Parasocial interaction
Parasocial interaction refers to a kind of psychological relationship experienced by an audience in their mediated encounters with performers in the mass media, particularly on television. Viewers or listeners come to consider media personalities as friends, despite having limited interactions with them. PSI is described as an illusionary experience, such that media audiences interact with personas as if they are engaged in a reciprocal relationship with them. The term was coined by Donald Horton and Richard Wohl in 1956.
Evolution of the term
Parasocial interaction was first described from the perspective of media and communication studies. In 1956, Horton and Wohl explored the different interactions between mass media users and media figures and determined the existence of a parasocial relationship, where the user acts as though they are involved in a typical social relationship. However, parasocial interaction has existed before mass media, when a person would establish a bond with political figures, gods or even spirits. Since then, the term has been adopted by psychologists in furthering their studies of the social relationships that emerge between consumers of mass media and the figures they see represented there. Initially, Horton and Wohl viewed parasocial interactions as abnormalities resulting from a lack of time spent with others. Perse and Rubin contested this view, finding that parasocial interactions occurred as a natural byproduct of time spent with media figures.Although originating from a psychological topic, extensive research of PSI has been performed in the area of mass communication with manifold results. Psychologists began to show their interest in the concept in the 1980s, and researchers began to develop the concept extensively within the field of communication science. Many important questions about social psychology were raised concerning the nature of these relationships that are problematic for existing theories in those fields. The concept of parasocial interaction and detailed examination of the behavioral phenomena that it seeks to explain have considerable potential for developing psychological theory.
Parasocial relationships
The concepts development of PSI and PSR are interpreted and employed in different ways in various literature. When it is applied in the use-and-gratifications approaches, the two concepts are typically treated interchangeably, with the regard primarily to a special type of "interpersonal involvement" with media figures that includes different phenomena such as interaction and identification. In contrast to the U&G approaches, research domains such as media psychology and semiotics argue for a clear distinction between the two terms. PSI specifically means the "one-sided process of media person perception during media exposure"; whereas PSR stands for "a cross-situational relationship that a viewer or user holds to a media person, which includes specific cognitive and affective components". Schmid & Klimmt further argue that PSI and PSR are progressive states such that what begins as a PSI has the potential to become a PSR. In sum, the terms, definitions, and models explicating PSI and PSR differ across scientific backgrounds and traditions.Scientific research
Studying social interaction, and by extension parasocial interaction, follows a social cognitive approach to defining individual cognitive activity. Accordingly, there are similar psychological processes at work in both parasocial relationships and face-to-face interactions. However, the parasocial relationship does not follow the process of the typical long-term relationship. The media user remains a stranger to the media figure, whereas this "strangeness" would gradually evaporate in typical social interaction. Many parasocial relationships fulfill the needs of typical social interaction, but potentially reward insecurity. Many who possess a dismissive attachment style to others may find the one-sided interaction to be preferable in lieu of dealing with others, while those who experience anxiety from typical interactions may find comfort in the lives of celebrities consistently being present. Additionally, whatever a celebrity or online figure may do can provoke emotional responses from their audiences—some even going as far as suffering from negative feelings because of it.The research of PSI obtained significant interest after the advent of the uses and gratifications approach to mass communication research in the early 1970s. In a study of early soap opera, McQuail et al. identified two essential functions of PSI: companionship and personal identity. Rosengren and Windahl further argued that PSI could be identified in the process of viewers' interacting with media figures, but such interaction did not produce identification. This is an important distinction, because identification has a longer history than PSI. Subsequent research has indicated that PSI is evident when identification is not present. During last several decades, PSI has been documented in the research analyzing the relationship between audience members and television newscasters, TV and radio talk-show hosts, sitcom characters and other TV celebrities or performers. Research has also been conducted on how a favorable PSI can be facilitated between celebrities and their followers on social media, specifically through the interactions followers have with the celebrities posts on social media. Although different PSI scales have been employed in these studies, PSI was clearly documented with each persona.
Noticing the importance of media in the area of psychological research, Giles asserted that there is a need for PSI research to move away from the field of mass communication and into the field of psychology. Studies in this area is commonly conducted by focusing a key psychological issue for PSI: concerning the similarity between parasocial relations and ordinary social relations. For example, Turner adopted the idea of homophily to examine the interpersonal and psychological predictors of parasocial interaction with television performers. The author found that one dimension of homophily was the best predictor of parasocial interaction. Hataway indicated that although there seems to be prevailing to analyze PSI in the domain of social psychology, a solid connection to psychological theory and developmental theory has been missing. Hataway further suggested that more psychological research is needed in order to develop parasocial theory. Specific issues cited were "how parasocial relationships are derived from parasocial interaction and the way those relationships further influence media usage as well as a social construction of reality, and how parasocial interaction is cognitively produced". He saw that the majority of PSI research has been conducted by mass communication scholars as a weakness and called for psychologists to refer to Giles for directions of studies.
Another important consideration for the study of PSI at a psychological level is that there is a form of PSI existing even in interpersonal social situation. People may use fundamentally the same cognitive processes in both interpersonal and mediated communication. Giles also suggested that the element of direct interaction occurred in mediated interaction, such as talking to a presenter or celebrity guest, may continue in social interaction, with a cartoon character or a fictional protagonist in the mind. This may finally constitute a new way of interpreting social interaction. A further consideration is application of social cognitive approaches in individual levels. It is traditionally accepted that this approach is inadequate by itself for the study of relationships. However, a number of growing literature on the role of imagination in social interaction suggests that some imaginative activity may be an influential factor in the outcome of real social interaction. PSI is nowadays regarded as an extension of normal social cognition, specifically in terms of the use of the imagination. Current PSI literature commonly acknowledge that the psychological processes acting at the individual level parallel those used in ordinary social activity and relationship building.
Psychological implications during childhood
Parasocial interaction is best explored across a lifespan, which explains the growing focus on parasocial interaction in children and adolescents. Studies have found that sex-role stereotyping is common in children's parasocial relationships with media figures, though boys most overwhelmingly choose male characters, while young girls are less likely to prefer one gender over another. Additionally, sex-role stereotyping is more common in children ages 5–6, but decreases in children age 10–11. Existing literature also intimates that attachments, parasocial or otherwise, established in early childhood, are highly influential on relationships created later in life. Many studies have focused on adolescent girls because they are more likely to form a strong bond with a media figure and be influenced in terms of lifestyle choices.Positive consequences
Identity formation
The primary effect is that of learning: consistent with Bandura's social cognitive theory, much evidence shows that children learn from positive and negative televised role models, and acquire norms and standards for conduct through media outlets such as television and video games. This is supported by a study by Cynthia Hoffner with children aged 7–12, which showed that the gender of children's favorite televised characters was strongly correlated to the gender of the children. Furthermore, the research showed "wishful identification" with parasocial relationships, namely, that boys preferred intelligence, while girls preferred attractiveness when picking favorite characters. These alternatives are both enhanced and mitigated by their separation from reality.On one hand, parasocial interactions are particularly appealing to adolescents in the throes of identity formation and increasing autonomy from parents because these relationships provide idealized figures with whom the adolescent can envision total acceptance. The lack of actual contact with these idealized figures can offer positive social interactions without risk of rejection or consequent feelings of unworthiness. One cannot know everything about a media figure or icon, allowing adolescents to attach fantasized attributes onto these figures in order to meet their own specific wants or needs. On the other hand, entities far removed from reality tend to be less influential on children.
A study by Rosaen and Dibble examined correlation between realism of favorite television character and strength of parasocial relationships. Results showed a positive correlation between social realism and strength of parasocial relationships. Results also show age-related differences among children. Older children tended to prefer more realistic characters, while younger children generally had more powerful parasocial relationships with any character. Age, however, did not impact the correlation between social realism and strength of parasocial interaction, which suggests that more real characters are grounds for more powerful parasocial relationships in children of all ages.
Learning through the media
The ability to learn from parasocial relationships is directly correlated to the strength of the relationship, as has been shown in work by Sandra L. Calvert and colleagues. In a study by Lauricella, Gola, and Calvert, eight 21-month-old infants were taught seriation sequencing by one of two characters. One character, Elmo, is iconic in American culture and therefore socially meaningful, and the other, DoDo, although popular with children in Taiwan, is less well known in American media. Children were better able to learn from the socially meaningful character than from the character who was less easily recognized. Furthermore, children could become better able to learn from less socially-relevant characters such as DoDo, by developing a parasocial relationship with that character. Accordingly, after children were given DoDo toys to play with, their ability to learn from that character increased. In a later study, this effect was found to be greatest when children showed stronger parasocial relationships: Children's success on the seriation task, and therefore their ability to learn from a less familiar character, was greatest for children who exhibited more emotional nurturing behaviors toward the DoDo toy during play.Personalization of a character makes a child more likely to nurture the character, and thus more likely to form a parasocial relationship that would improve learning from videos featuring the character. In place of DoDo and Elmo, Calvert et al. gave children Scout and Violet dolls. These interactive plush toy dogs can be programmed to say a child's name and have particular favorites. 18-month-old children were given either personalized toys or non-personalized toys. At the end of the study, children who had received personalized dolls were better able to learn from their characters than were children who had received non-personalized toys. Children also nurtured personalized toys more than non-personalized toys. It seems that perceived similarities increase children's interest and investment in the characters, which motivates the development of parasocial relationships and helps improve later screen-based learning.
Negative consequences
In the past two decades, people have become increasingly interested in the potential negative impacts media has on people's' behavior and cognition. Many researchers have begun to look more closely at how people's relationships with various media outlets affect behavior, self-perception and attachment styles, and specifically in regards to creating parasocial relationships.Body image
Further research has examined these relationships with regard to body image and self-perception. Interest in this more narrow area of research has increased as body image issues have become more prevalent in today's society.A study was conducted to examine the relationship between media exposure and adolescents' body image. Specifically, researchers looked at parasocial relationships and the different motivations for self-comparison with a character. This study surveyed 391 7th and 8th grade students and found that media exposure negatively predicted body image. In addition to the direct negative impact, the study indicated that parasocial relationships with favorite characters, motivations to self-compare, and engagement in social comparison with characters amplified the negative effects on kids' body images. Furthermore, the researchers found that making social comparisons with favorite characters distorted actual, or ideal, body image and self-perception. Studies have been done exploring these effects across gender.
A study examined the parasocial relationships between men and superheroes; the study looked at muscular vs. non-muscular superheroes and men who either did or did not develop a one-sided psychological bond with a superhero character. The results from this study indicated a significant impact on body image, particularly when exposed to muscular superhero characters. Research conducted by Ariana F. Young, Shira Gabriel, and Jordan L. Hollar in 2013 showed that men who did not form a parasocial relationship with a muscular superhero had poor self-perception and felt negative about their bodies after exposure to the muscular character. However, if the men had a PRS with the superhero, the negative effects on body satisfaction were eliminated.
Aggression
Further studies have looked into parasocial relationships and more specifically at the impacts on violent and aggressive behavior. A study done by Keren Eyal and Alan M. Rubin examined aggressive and violent television characters and the potential negative impacts they may have on viewers. The study was based on social cognitive theory and looked at trait aggression in viewers and identification and parasocial interaction with aggressive characters. The researchers measured trait aggression in each of the participants and compared that to the level of identification with aggressive characters. The study found that more aggressive viewers were more likely to identify with aggressive characters and further develop parasocial relationships with the aggressive characters.Parasocial interaction has been linked to psychological attachment theory and its consequences have seen the same dramatic effects as real Relationship breakup. In considering the relationship between parasocial interaction and attachment styles, Cohen found that individuals who were more anxious media consumers tended to be more invested in parasocial relationships.
In parasocial interaction there is no "normal" social interaction; it is a very one-sided relation. The knowledgeable side has no direct control over the actions of the side it observes, and it is very difficult for it to contact and influence it.
Parasocial breakup
While much research focuses on the formation and maintenance of parasocial relationships, other research has begun to focus on what happens when a parasocial relationship is dissolved. Eyal and Cohen define parasocial breakup as "a situation where a character with whom a viewer has developed a PSR goes off the air." The distress that media consumers experienced after a parasocial breakup was quite similar to that of a social relationship. However, the emotional distress experienced after the parasocial breakup was weaker than that of the real life interpersonal relationship. Lather and Moyer-Guse also considered the concept of parasocial breakup, but in a more temporary sense. While the study focused on parasocial breakups as a result of the writers' strike from 2007–2008, the researchers found that media consumers still experienced different levels of emotional distress. This study, like previous studies, showed that parasocial relationships operate very similarly to real-life relationships.On the internet
In 1998, John Eighmey, from Iowa State University, and Lola McCord, from the University of Alabama, published a study titled "Adding Value in the Information Age: Uses and Gratifications of Sites on the World Wide Web." In the study, they observed that the presence of parasocial relationships constituted an important determinant of website visitation rates. "It appears," the study states, "that websites projecting a strong sense of personality may also encourage the development of a kind of parasocial relationship with website visitors."In 1999, John Hoerner, from the University of Alabama, published a study titled "Scaling the Web: A Parasocial Interaction Scale for World Wide Web Sites", in which he proposed a method for measuring the effects of parasocial interaction on the Internet. The study explained that websites may feature "personae" that host to the visitors to the sites in order to generate public interest. Personae, in some cases, are nothing more than the online representations of the actual people, often prominent public figures, but sometimes, according to the study, will be the fictional creations of the sites' webmasters. Personae "take on many of the characteristics of a companion, including regular and frequent appearances, a sense of immediacy...and the feeling of a face-to-face meeting." Additionally, the study makes the point that, even when no such personae have been created, parasocial relationships might still develop. Webmasters might foster parasocial interactions through a conversational writing style, extensive character development and opportunities for email exchange with the website's persona.
Hoerner used the Parasocial Interaction scale, developed by Rubin, Perse, and Powell in 1985, and modified the scale to more accurately assess parasocial interactions on the Internet. They used the scale to gauge participants' reactions to a number of different websites, and, more generally, to determine whether or not parasocial interaction theory could be linked to Internet use. The study concluded, first, that parasocial interaction is not dependent on the presence of a traditional persona on a website; data showed that websites with described "strong personae" did not attract significantly more hits than other websites selected by the study conductors. "The literal, mediated personality from the newscast or soap opera of the past is gone. The design metaphor, flow of the web experience, and styles of textual and graphic presentations of the information all become elements of a website persona and encourage parasocial interaction by the visitor/user with that persona."
In social media
Though most literature has focused on parasocial interaction as a television and film phenomenon, new technologies, namely the Internet, have necessitated a closer look at such interactions. The applications of PSI to computer-mediated environments are continuously documented in literature in last decade. Many researchers concluded that, just as parasocial relationships are present in television and radio, they are also present in online environments such as blogs and other social networking sites. Through an exploration of followers on politicians' blogs, Thorson and Rodgers found that parasocial interacting with the politician influences people's opinions about the politician, and promotes them to vote for the politician. Social media is designed to be a new channel through which parasocial interaction/relationship can be formed. Research has shown that interacting with individuals through blogs and social media such as Twitter can influence the perceptions of those individuals. As Internet users become more active on social media platforms such as Facebook and Twitter, followers often feel more engaged with them, making the parasocial relationships stronger.Social media is defined as "Internet-based applications that build on the ideological and technological foundations of Web 2.0 and that allow the creation and exchange of user-generated content". While the usage of social media for personal means is common, the use of social media by celebrities has given them an opportunity to have a larger platform for personal causes or brand promotion by facilitating word-of-mouth.
Social media networks inherit at least one key attribute from the Internet, in that they offer open accessibility for all users. Philip Drake and Andy Miah argue that the Internet, and therefore social networks and blogs, downsize the gatekeeping processes that exist in other mass media forms. They further state that this means that online information can spread unfiltered and thus does not rest on strict framework conditions such as those on television or in newspapers. This, however, remains subject to an ongoing debate within research. Through presence on social media platforms, stars and celebrities attempt on the one hand to participate in the production of their image; on the other hand, they must remain present in these media in order to stay on the media's and consequently on the audience's agenda. According to Daschmann, celebrities all have to compete for the public's attention. In such a competitive environment a famous person must therefore remain present on all the accessible media channels., a Let's Play content creator with over 22 million subscribers
YouTube
, a social media platform dedicated to sharing video-related content produced by its users, has grown in popularity to become a form of media that's likened to television for the current generation. By content creators granting insight into their daily lives through the practice of vlogging, viewers form close one-sided relationships with these creators that manifest in comment chains, fan art and consistent responses with the creator in question. Parasocial interaction and relationships are commonly formed between the creators and their audiences due to the creator's desire to interact with their fanbase through comments or posts. Many creators share "personal" details of their lives, even if there's little authenticity in the polished identity they convey online. In a study conducted by Google in 2017, a reported 40% of millennial YouTube subscribers claimed their "favorite creators understands them better than their friends." For many viewers, parasocial relationships check off the four factors that are defined by Mark Granovetter's "The Strength of Weak Ties" theory: intimacy is gained by the creator's sharing of personal details, by which their viewers may react emotionally; viewers dedicate time to watching content the creator uploads; and what the creator posts—whether sponsored or not—may make the viewer feel as if they are being offered something, like a favor.Commercial influences
Parasocial interaction theory was used to understand consumers' purchasing behavior in online context. With the development of social media, such as Facebook, Twitter, Instagram, and YouTube, both companies and consumers start to use social commerce platforms more frequently. Many studies indicate that, among various factors affecting consumers' purchase decision on SCPs, such as credibility of products, parasocial interaction exerts more potential influence on users' final decisions. Sokolova and Kefi conducted a study with large data set from the audiences of four popular influencers in the beauty and fashion sector in France to discover the influence of parasocial interaction and credibility on consumers' purchase intention. Their study discovers that younger generations value parasocial interaction and their personal attachment to influencers more than credibility.On the social commerce platforms, users intend to build parasocial interaction, one with other users, one with celebrities.
PSI with other users
Certain social media users are active creators of online content, such as personal experiences, ideas, reviews, for targeted audience, which are called influencers. Influencers can become experts, similar to celebrities to some extent, and their posts may influence products and brands and affect potential customers, i.e. their followers. The users in a social commerce platform "meet" with other users and influencers through the images, videos, and feedbacks that they share on the social media. By the time, after multiples times of "meetings", the imaginary intimacy is improved, and the users will deliberately maintain the online friendship, which is a parasocial interaction. Influencers on the social media platforms often comment on the products they have tested, and promote them online to other users by providing their feelings and self-experiences along with images and videos. Under the parasocial relationship, users intend to rely on the images or comments that influencers had on the products, which will influence consumers' final decisions. Thus, many social commerce merchants utilize this psychological implication, and sift beautiful images and positive comments on products to provide users a more intuitive shopping experience. For instance, the occurrence and popularity of vlog creates a social space where strangers can share feelings and build intimate relationship. Many merchants pay vlogger for recommending and persuading their subscribers to purchase the products.PSI with celebrities
PSI relationships are more readily formed between social media user and celebrities. On social media, celebrities build and strengthen more intimate relationships with consumers and fans. Simultaneously, the context of SCPs, supported by Web 2.0 social media technologies, stimulates users' parasocial interactions with celebrities and experts. Uncertainty reduction theory is an example of a way that this can occur. The process of repeated exposure to an individual gradually reduces the user's level of uncertainty, which increases the user's chances of liking this celebrity. Repeated exposure to the celebrity gives users a sense of predictability in their actions, which engenders a sense of loyalty. Users who are immersed in celebrity-fans PSI may affirm their loyalty through various activities, including purchasing products endorsed by celebrities. Unlike influencers, celebrities bring their fans with stronger impulse purchase. Targeted consumers desire to interact with celebrities, instead of passively receiving information from celebrities. By purchasing and supporting the celebrity endorsing products, fans may build more intimate relationship with celebrities in their imagination. In a 2014 journal article, Seung-A Annie Jin and Joe Phua discussed how they conducted studies to determine multiple hypothesis based on the number of followers a celebrity had in correlation to the trust that imparted onto a consumer. This study was done in terms of a celebrity endorsing a product and the likelihood of the consumer to purchase the product after seeing the promotion. Consumers perceived the celebrity with a high number of followers as being more physically attractive, trustworthy, and competent. A high number of followers on the celebrity endorser's profile also significantly increased consumers' intention to build an online friendship with the celebrity. The study found that if a celebrity with a higher number of followers was perceived as more trustworthy, the consumer exhibited significantly higher postexposure product involvement and buying intention as opposed to those who were exposed to a celebrity with a lower follower count. Merchants on social commerce platforms will find huge potentials of analyzing and applying parasocial interactions to manipulate consumers' purchase intention.PSI with companies
As social media relationships grew between celebrities and influencers, businesses created social media profiles for audience engagement. Fast food restaurants have started comedy twitter accounts to interact with their customers in a personal way. The companies' twitter accounts respond to tweets from customers, tell jokes and engage in the online industry in ways that create PSI with the consumers. This strategy is working. A study done by Lauren I.Labrecque in 2013 found that customers have higher loyalty intentions and are more likely to provide information for the brand when the brand fostered PSIs. The study also showed that these outcomes were less likely when the consumer felt the response from the company's social media account were automated. Furthermore, including personal details and behind-the-scenes ideation in interactions with consumers also triggers PSI and results in a positive impact.Limitations
Most studies find that PSI only occurs as friendship, which is overly restrictive theoretically and practically. It is common that people may build parasocial interactions with certain media figures, even though they do not consider to be "friends", such as a villain in a show. Though PSI with disliked figures occurs less likely than with heroes and positive characters, the situation of "love to hate" relationship with disliked characters still occurs. Some researchers realize the restriction of limiting PSIs as friendship, which may preclude them from capture broader situation of meaningful media user reactions. In 2010, Tian and Hoffner conduct an online questionnaire measuring the responses from 174 participants to a liked, neutral, or disliked character from the ABC drama Lost. All participants reported the identification they had perceived with the character, as well as the parasocial interaction and how did they try to change their perspectives to be more like the character. According to the whole sample, perceived similarity was a significant positive predictor of both identification and parasocial interaction. Undeniably, parasocial interaction was higher for liked than for neutral and disliked characters based on the study. Parasocial interaction still appeared with liked, neutral and disliked characters. The prevailing perspective of PSI as a friendship is not appropriate based on the theoretical and experimental findings, and many researchers start to improve the measurement of the PSI's concept.Future research
One direction for future PSI concerns the advancement of methodology. As theories become more defined and complex, experiments seem to be necessary to be employed in testing hypotheses. Because the meanings of perception and emotion take up much of what parasocial interaction/relationship research interest, the cause and effect is hard to be distinguished and potential spuriousness is difficult to be avoided. For example, whether similarity precedes PSI and whether mediated interaction create a sense of similarity requires experimental validation.Cohen also suggested that different types of relationships are encouraged to be analyzed by different genres, which particularly challenges scholars in examine the mediated relationship in those reality TVs. These prototypical reality shows which are built around narratives, displaying a lot of emotions which seem to invite empathy and identification, and also demonstrating the characters' skills as to develop fandom. Ratings and audience responses provide strong evidence that those reality shows create significant mediated relationship, but future inquiries should examine whether this new kind of mediated interaction/relationship evolves or do these interactions/relationships conform to existing patterns.
The influence of media in childhood has received little attention from developmental psychologists, even though children have a high degree of exposure to media. While many studies and experiments have explored the nature of parasocial relationships, there are many opportunities for future research. For example, a potential future area of research could be the issue of reruns, where the relationships have outcomes which are already known or well-established. In addition, another area of research could focus on production techniques or televisual approaches. This would include techniques such as chiaroscuro or flat lighting, the strategic placement of close-ups or establishing shots, deductive or inductive shot sequences, hip hop editing, or desaturation. These techniques have long been theorized to have some sort of influence on the formation of parasocial relationships, but their influence has yet to be determined.
The prevailing use of social media and its impact on mediated relationships also requires further study of PSI. Different social media platforms provide channels through which celebrities communicate with their followers easily, making parasocial interaction/relationships seem less unidirectional and perhaps more satisfying and intense. As such, whether social media has made PSI more a part of everyday life needs further exploration. Technological development has been raising questions regarding the role of PSI in our social lives, as media content is available in more places and times. Our mediated friends are never too far away, instead, they actually rest in our pockets and sleep in our beds. Whether this means that we spend more time and effort on cultivating these relationships and will be less dependent on real social relationships, needs further exploration.
Other concerns include the continuity of media figures representation across various media outlets, and the notion of parasocial interaction as compensation for lack of social outlets. Popstars, for example, may not only appear on television, but on several different television or radio programs, as either a chat guest or a performer; further repeated viewings of these stars would intensify visual aspects of parasocial interaction with that star. Most research has typically characterized media users as a television viewer who is often solitary and in need of social interaction. The different types of user-figure interaction can be addressed by conceptualizing parasocial interaction as an extension of ordinary social interaction. Through close examination of social encounters that are significant for parasocial relationships, we can continue to distinguish between parasocial interaction an isolated activity and longer-term interaction.