Membership categorization analysis as means of studying person perception

ABSTRACT Membership categorization analysis focuses on studying how people categorize each other in interaction. Even though the role of observation has been acknowledged within this methodological approach, it has mainly been used to analyze oral and written communication. This study focuses on examining how membership categorization analysis can be applied to person perception. It elucidates the procedure with two examples where students in a vocational upper secondary college in Finland perceived portrait paintings and reflected in writing what kind of teachers the people depicted in them would be. Drawing on the theorization of person perception and membership categorization analysis, the study suggests a model for applying membership categorization analysis in person perception and discusses its relevance to social interaction.


Introduction
The visual plays a crucial role in our everyday lives. Not only do we live, work and act in visual environments and premises, but we also encounter people, objects and phenomena that are visual. In addition, we are surrounded be a plethora of images in diverse forms of media. Instead of being lost in the visual, we are capable of navigating our lives in these visual circumstances.
The central precondition for coping with the visual is that we do not assess an object's visual qualities without assigning meaning; instead, we perceive them as meaningful. As members of cultures and societies, we do not randomly link a visual form to any meaning crossing our minds but adhere to the situationally relevant aspects of shared cultural experience and knowledge (social categories) (Kress and van Leeuwen 2006;Sturken and Cartwright 2017). In social situations, we do not employ social categories as rigid and stable acontextual templates. In contrast, they become contextually redefined in everyday categorizing practices (Watson 2015). Hence, meaning construction regarding visual entities that is performed by drawing on social categories is not a mechanical transference of cultural meanings onto visual encounters. Instead, the meaning construction itself determines how the cultural resources are applied (e.g., Francis and Hester 2017;Watson 2015).
Clusters of visual forms together with their cultural meanings form visual orders that can be related to socially constructed categories concerning visuals in our everyday lives (Seppänen 2006). Visual orders form the resources that enable us to make sense of the visual encounters and to communicate with each other meaningfully using visual means (Lepper 2000;Seppänen 2006). Similar to social categories, the use of visual orders is situational, and the situation determines which aspects of the visual orders are selected as the resource of sense making (Seppänen 2006).
Harvey Sacks developed membership categorization analysis in the 1960s to study how people categorize each other in social interaction (Fitzgerald 2015;Housley and Fitzgerald 2015;Lepper 2000). Membership categories are resources for meaning-making and in its basic form, membership categorization means identifying a person as a member of a social category (Freiberg and Freebody 2009;Järviluoma and Roivainen 2003;Paulsen 2018). Even though Sacks (1972bSacks ( , 1992 recognized and acknowledged the role of visual perception in membership categorization from the very beginning, and there are a number of studies using in situ perceptions as well as visual and multimodal materials in this tradition, membership categorization analysis has predominantly been applied to verbal accounts (Fitzgerald 2015;Francis and Hart 1997). This particular study focuses on examining membership categorization analysis related to person perception. More precisely, this study asks how people categorize each other based on visual cues.
With regard to the aforementioned research task, this study first presents and discusses person perception and membership categorization analysis. Thereafter, it elucidates the possibilities of membership categorization analysis to capture the categorization in person perception through examples of empirical data where students in upper secondary vocational education in Finland perceived portrait paintings as images of teachers and categorized them based on visual cues. Based on the findings, the study suggests a model for applying membership categorization analysis to person perception.

Person perception
In social psychology, there are a number of approaches to person perception. For instance, traditionally, scholars have focused on the causal explanations of attribution theory when observing the behaviors of other people. More recently, a number of scholars have integrated perceptual interpretation and categorization processes to account for top-down and bottom-up dynamics within social reasoning. (Freeman and Ambady 2011;Kuzmanovic et al. 2013) This study adheres to the latter approach to person perception with the aim of showing how students categorize teachers based on visually perceivable features when perceiving them in images.
People use visual nonverbal cues of other people as sources of information on the basis of which they draw inferences about their thoughts, feelings, characters, traits, motivations and behavioral intentions (Hess, Adams, and Kleck 2008;Skowronski and Ambady 2008). Skowronski and Ambady (2008) regard the urge to understand and explain these directly unobservable "matters' through what is observable as a basic human motivation in order to predict other persons' behavior and the course of future interaction. Even though these judgements and impressions are often based on momentarily perceived features, the perceivers tend to regard them as relatively enduring dispositions (Zebrowitz and Montepare 2008). Among the physical appearance cues, facial expressions play an important role since they are regarded as the major means of expressing emotions and revealing information about people's intentions (Skowronski and Ambady 2008;Zebrowitz and Montepare 2008).
We categorize people, events and actions constantly. In person perception research, the term 'category' is used to refer to the totality of information that perceivers have in mind about various groups of individuals (Macrae and Bodenhausen 2001;Mason, Cloutier, and Macrae 2006). Social categorization is regarded as fundamental to social perception, thought and action, since it organizes and structures the social world as well as facilitates our communication about it (Augoustinos, Walker, and Donaghue 2014). Classifying people based on their social roles -e.g. occupation -is one form of social categorization providing people with normative expectations about the behavior of individuals occupying these roles (Augoustinos, Walker, and Donaghue 2014).
The conception that categorization plays an important role in person perception has a long tradition and can be traced to seminal writings of Bartlett (1932), Allport (1954) and Tajfel (1969), according to whom categorical conceptions based on prior knowledge and experiences about the social world dominate the processes of perception and interpretation of visual cues and occur mostly automatically. Categorization is still understood to play a central role in person perception. However, its automaticity and inevitability have been challenged by, for example, dual process theories of impression formation, such as the continuum model (Fiske and Neuberg 1990) and dual process model (Brewer 1988), that regard the perceivers as possessing multiple cognitive strategies and being able to choose between them depending on the level of their involvement with the target. In other words, perceivers move from category-based observations to more individuated and piecemeal-constructed impressions depending on their motives, attitudes, and goals toward the target (Gawronski and Creighton 2013;Macrae and Bodenhausen 2001;Mason, Cloutier, and Macrae 2006). These more nuanced and flexible approaches of social categorization have gained support in recent theorizing (Augoustinos, Walker, and Donaghue 2014).

Membership categorization analysis
Categorization is a basic human activity (Eskelinen, Olesen, and Caswell 2010;Hogg 2003). Bowker and Star (2000) suggest that categorization is often unconscious and directs people's actions tacitly (see also Fitzgerald 2015). In many cases, our habitual ways of categorization become visible when something deviates from our expectations (Bowker and Star 2000;Sacks 1972b). Categories are often considered as stores of cultural knowledge and constituents of social order facilitating social interaction and rendering it predictable (D'hondt 2013;Sacks 1992;Tanner 2019). According to Stokoe (2012, 37), 'categories short-cut and package common sense knowledge about category members and their actions.' However, referring to Stephen Hester's work, Francis and Hester (2017) warn about the 'problem of culturalism,' having noted that in membership categorization analysis, cultural categories cannot be understood as rigid and unchanging templates external to the categorization. In contrast, culture is constituted in categorizing practices: 'culture (. . .) exist in action' (Hester & Eglin1997a, p. 20). Hence, cultural categories are occasioned constructions that are actively and selectively applied and modified in occasioned categorizing practices (Francis and Hester 2017, 56).
As previously suggested, categories are understood as structures that are under a process of constant reconstruction: when people draw on culturally constructed categories to make sense of social encounters, the categories do not remain unaltered but are subject to modification (Fitzgerald 2015;Hester and Eglin 1997a;Silverman 2006). Thus, categorization can be understood as the construction of social realities and order (Housley and Fitzgerald 2015;Lee 2018;Tanner 2019). When categorizing people, the choice of categories expresses what we regard as normal, valuable or inappropriate for an incumbent of the category (Martikainen 2017;Jayyusi 1984). According to Sacks (1972b) these kinds of judgments are also made when we categorize people based on their appearance.
Membership categorization analysis originated from the work of Harvey Sacks in the 1960s, which focused on examining how people as members of various social communities locally categorize people and their actions on the basis of their mundane, common sense reasoning (Fitzgerald 2015;Housley and Fitzgerald 2015;Paulsen 2018). He was interested in examining the principles and methods that members use when they categorize themselves as well as other people into various membership categories in social situations (Sacks 1972a(Sacks , 1986(Sacks , 1992. Besides categories and methods, the interactional consequences of the categorization were at the core of Sacks's interest (Fitzgerald 2015;Hester and Eglin 1997a). For this purpose and in congruence with ethnomethodological tradition, the method focused on analyzing naturally occurring data -such as situational speech -for studying social interaction (Fitzgerald 2015;Garfinkel and Sacks 1986;Lee 2018;Ylänne and Nikander 2019). Since Sacks, membership categorization analysis has also been used to analyze non-personal objects and entities (Freiberg and Freebody 2009;Housley and Fitzgerald 2015).
Membership categories -and the rules governing their application -are not cognitive prototypes inside people's heads or inflexible conceptual grids which people automatically impose when categorizing (D'hondt 2013). Instead, they are socially constructed and commonly shared resources that help people navigate social situations (Fitzgerald, Housley, and Rintel 2017;Järviluoma and Roivainen 2003). In everyday social interaction, categorization is a routine activity based on common sense reasoning with the aim of describing, explaining and predicting other people's behavior (Kärkkäinen et al. 2019).
According to Sacks (1972bSacks ( , 1992, membership categories can be defined as classifications or social types that are used as means of describing people and making sense of their local, situational actions (see also Hester and Eglin 1997a;Paulsen 2018). Hence, membership categorization is understood as a procedure in which cultural and social constructions are selectively intertwined with situational factors (Fitzgerald, Housley, and Rintel 2017;McLay & Renshaw 2019). This kind of navigation between cultural resources and situational encounters that occurs in purposeful action has been conceptualized as 'culture-in-action' (Hester and Eglin 1997b, 28) and 'social-knowledge-in-action' (Fitzgerald, Housley, and Rintel 2017, 51). Jayyusi (1984), for her part, conceptualizes the categorical work as the interplay between culturally available membership categories and more situational membership categorizations. As a site for the study of context and culture, membership categorization analysis intertwines macro and micro levels of analysis (Hester and Eglin 1997c).
Membership categorization is not a neutral activity, since membership categories not only describe and classify people but also make normative assumptions and judgments about people's identities (Baker 1997;Lee 2018;Raymond 2019;Ylänne and Nikander 2019). In this sense, categorization constructs identities and social norms and strongly influences social interaction (Bowker and Star 2000;Housley and Fitzgerald 2015;Lee 2018). As Housley and Fitzgerald (2015, 2-3) point out, membership categorization analysis gives an approach to examine social identity as something 'people in society do, achieve, negotiate, attribute things to and act upon as part of their daily lives.' From the perspective of membership categorization analysis, all identities are dynamic and examining the categories people produce in their everyday activities is a means of studying them (Lee 2018;McLay & Renshaw 2019;Ylänne and Nikander 2019). Housley and Fitzgerald (2015) argue that Sacks's interest didn't focus so much on mere explication of social categories but on the way people combine categories with descriptive expressions in particular situations in order to define the content of categories contextually. Similarly, Cuff (1993) suggests that 'general categories' often appear to be too vague and abstract for the purposes of interaction and must be refined with descriptions that characterize more adequately members' situational actions in a particular category (see also Hester and Eglin 1997a;McLay & Renshaw 2019). Sacks (1992) himself paid attention to category-bound activities that characterize actions which people, as incumbents of a category, undertake.
Additionally, scholars have paid attention to 'features' (Jayyusi 1984), 'predicates' (Watson 1997), and 'attributes' (Fitzgerald 2015;Housley and Fitzgerald 2015) that express category-bound norms and expectations concerning people acting in a particular membership category. In accordance with the aforementioned scholars, Jayyusi (1984) argues that situationally constructed membership categorizations often follow the model 'adjective plus category-concept,' where the adjective contextually specifies the culturally available membership category.
We do not categorize only verbally, but also visually (Bowker and Star 2000;Lepper 2000;Sacks 1992). Even though membership categorization analysis has predominantly been applied to analyzing verbal accounts (Fitzgerald 2015;Francis and Hart 1997), Sacks (1972bSacks ( , 1992 recognized and acknowledged the role of visual perception in membership categorization from the very beginning, arguing that people categorize each other based on their looks. In his two "viewer's maxims", Sacks (1992) conceptualized person perception as an activity in which cultural norms and situated perceptions of people interlace forming membership categories (see also Paulsen 2018;Sacks 1986). He stated that categorization serves as a means and source of drawing inferences about the persons included in the category. These inferences may relate to more stable personality traits as well as situational thoughts, emotions, actions, and intentions (Sacks 1972b(Sacks , 1992. Thus, studying person perception may provide means of detecting visual orders and cultural norms embedded in social interaction, contributing to our understanding of social reality (Housley and Fitzgerald 2015;Sacks 1992;Watson 1997).
Observations of people's behavior and interaction in public places as well as photos and videos capturing situations of everyday interaction has functioned as rich material for membership categorization analysis, not only for Sacks (1972aSacks ( , 1972bSacks ( , 1986Sacks ( , 1992 but also for other researchers (Ball and Smith 2011;Carlin 2003;Hester and Francis 2003;Watson 2005). Recent research has become increasingly interested in using visual materials as research data. Among others, Francis and Hart (1997), Furukawa (2012), Raudaskoski (2011), Roca-Cuberes and Ventura (2016), and Stokoe (2012) have utilized TV programs, news, and commercials in research on membership categorization. Videos have been used, for instance, by Burdelski and Fukuka (2019), Edwards and Fitzgerald (2015), Evans and Fitzgerald (2017), Licoppe (2015), Reynolds (2017), Rintel (2015), Smith (2017), and Watson (2005). In addition, Stokoe and Attenborough's (2015) study used social media and Hall, Gough, Seymor-Smith, and Hansen's (2012) study utilized magazine photographs as the data. Verbal comments based on the perception of portrait paintings have served as data for Martikainen (2017Martikainen ( , 2020, Martikainen and Hujala (2017), and Martikainen, Laulainen, and Hujala (2019). Similar to them, several studies have used multimodal data combining visual and verbal materials (Housley and Fitzgerald 2015).

Participants and data
The examples included in this study are part of the data collected for a larger study where 65 students at an upper secondary vocational college for culture studies in Finland perceived and categorized portrait paintings as images of teachers (see Martikainen, 2020). The data were collected in the classroom situation in the college in question during the lesson on visual culture studies. The students were presented with 17 images depicting people with diverse visual features (facial expressions, gestures, posture, clothing, hair style, makeup, and location in the picture space). The students were asked to observe the images as visual representations of teachers and comment in writing on what kind of teachers the people would be. In reality, the people depicted in the paintings were not teachers. For this reason, the word 'teacher' is written in quotation marks when referring to persons depicted in the paintings. Perceiving people in images differs from perceiving people in real-life situations. However, several studies have found people use similar strategies (e.g., categorization) and pay attention to similar features when perceiving people in images and in real-life encounters (Brooks and Freeman 2018;Oosterhof and Todorov 2008). In addition, in both situations, people use their knowledge and experience of social life when categorizing the people (Demarrais and Robb 2013;Yang 2015).
The aim of this study is neither to present the analysis of the whole data nor the findings related to them. Instead, the aim of this study is to exemplify, through two image examples and written reflections related to them, how membership categorization analysis can be applied when studying person perception. Below, two images of the paintings and examples of students' verbal comments will be presented.
In this study, the purpose of the verbal comments is to elucidate the categorizing practices of the students when perceiving the 'teachers' in the images -namely, which visual features the students use as bases for categorizing the 'teachers' and what kinds of categories of teachers (membership categorizations) they construct based on the visual features. The students were told in advance to perceive the images as if they were images of teachers. Following Jayyusi (1984), this study understands 'Teacher' as the culturally available membership category and the ways students categorize 'teachers' in the images as membership categorizations. Even though there is a culturally constructed understanding of the basic determinants of being a teacher in Finland, what it means to be a teacher, and what is expected from a teacher differ from situation to situation (e.g., at different levels of education). However, in this study, the culturally available membership category 'Teacher' was not specified. In this study, the construction of membership categorizations is understood as an occasioned operation where the culturally available membership category 'Teacher' is situationally specified in the process of categorization. The visual features, in turn, are supposed to act as 'perceptually available category features' (Jayyusi 1984, 73) based on which the students form situational membership categorizations in terms of 'a certain kind of teacher'.

Findings
The findings present two images of teachers and membership categorizations related to them. The category label follows Jayyusi's (1984) adjective plus category-concept model in which the adjective situationally specifies the culturally available membership category.

'Friendly teacher', 'fair teacher', 'absent-minded teacher'
The first example of membership categorization is based on the students' perception of the Self-Portrait painted by Elin Danielson-Gambogi in 1889.
The majority of the students categorized the 'teacher' in the first painting as a 'Friendly teacher'. This way of categorizing was based on several visual features. Students used several visually perceivable category features as the basis of categorization. While some students paid attention only to the facial expression that they characterized as friendly, other students associated the white color of the dress, round face, and curly hair with friendliness. Membership categorization seemed to have interactional implications, since students assumed this kind of teacher would prepare her lessons well and put effort into teaching. In addition, students thought it would be easy to communicate with her. This is a kind and friendly teacher, who loves teaching. All the pupils would like her. White dress, smile and childish round face communicate friendliness and kindness. She is committed to her work and wants that everybody learns. (Student 15) She would be a friendly teacher. Warm colors, curly hair and calm facial expression make me think so. She would support her students and help them to learn. (Student 22) A number of students paid attention to the natural overall appearance of the 'teacher' in this category. The naturalness was communicated through neutral facial expression, straight gaze, neutral clothing style and lack of make-up. Based on these features, the 'teacher' was also categorized as a 'Fair and reliable teacher' who treats her students equally and is ready to listen to the problems the students have. She would be a fair teacher who treats everybody equally. Her straight gaze, serious but neutral facial expression, neutral clothing style and natural overall appearance make her look honest. It would be easy to talk to her. (Student 45) The face is relaxed, and there is the impression of honesty due to the lack of make-up. A relaxed, confident and reliable teacher. The kind of person you could take your problems to, without being judged. (Student 49) Some students perceived the 'teacher's' gaze as empty and facial expression as neutral or expressionless. In addition, blown hair and simple dress were perceived as communicating casualness. Based on these visually perceivable category features, a few students categorized this 'teacher' as an 'Absent-minded and indifferent teacher'. They thought absentmindedness would result in inconsistent and disorganized teaching. In addition, some students assumed this 'teacher' would have problems in maintaining order in the classroom. This teacher looks melancholy maybe even depressed. Because of this, she would be indifferent as a teacher. Her empty gaze and neutral facial expression communicate those meanings. Perhaps, she would be absent-minded and not consistent in her teaching. (Student 41) She looks friendly but absent-minded. Her hollow gaze, blown hair and simple dress create the impression of casualty. She would be a relaxed teacher who doesn't care about the rules and timetables. She would forget things and students should remind her of them. (Student 63) Perceiving the visual features of the person depicted in the painting not only generated visual associations but also associations related to sound and movement. For instance, one student thought the voice of the teacher would sound sleepy and she would move slowly.
Her voice would sound sleepy and she would move very slowly. Even though she looks friendly, I get the impression that she is gloomy or sad. She looks very sensitive and I doubt if she can maintain discipline in the classroom (Student 54) In sum, most students categorized the person in the painting as a 'Friendly teacher'. For them, the kind facial expression, open gaze, round face, curly hair, and white dress of the person served as visually perceivable category features (Jayyusi 1984), based on which the membership category 'Teacher' was specified contextually as a 'Friendly teacher'. This situationally constructed membership categorization (Jayyusi 1984) was associated with a teacher who is approachable, communicative, and committed to counseling her students. Hence, the visual features of 'Friendly teacher' generated expectations in terms of category-bound activities (Sacks 1992); 'Friendly teacher' was expected to be one who prepares her lessons well and willingly communicates with students. Based on the same aforementioned visual cues, some students categorized the person as a 'Fair and reliable teacher'. They expected this kind of teacher to treat her students equally and be willing to help students with their problems.
However, some students categorized the same person as an 'Absentminded teacher' or an 'Indifferent teacher'. They perceived the person's gaze as empty, facial expression as expressionless or melancholy, and hair style and dress as casual. These features acted for them as visually perceivable category features based on which the membership categorizations 'Absent-minded teacher' and 'Indifferent teacher' were constructed. These membership categorizations generated category-bound expectations that the person would not be capable of teaching in a consistent manner and keeping order in the classroom.

'Strict teacher', 'tired teacher', 'irresponsible teacher'
The second example of membership categorization is based on students' perceptions of Tyko Sallinen's painting Self-Portrait from 1914.
Several students categorized the man in the second painting as a 'Strict teacher'. The facial expression perceived as strict, cruel, angry, or harsh formed the most important basis of categorizing the man as a 'Strict teacher'. In addition, some students perceived the cruel gaze and tight or tense mouth as contributing to the air of strictness. Membership categorization generated feelings of fear among students. In addition, they thought that it would be difficult to communicate with this kind of teacher. This man has a very intensive and even cruel gaze. The facial expression is angry. For this reason, he would be a determined and strict teacher whom the students would be afraid of. (Student 25) The expression on his face is strict and judgmental. This teacher is strict and he punishes his students if they make any mistakes. (Student 8) Rude facial expression, tight lips and cold colors make me think that it would be difficult to approach this teacher. He keeps strict discipline and students would be afraid of him. (Student 46) Instead of a 'Strict teacher', some students categorized the 'teacher' in the second painting as a 'Tired teacher'. For them, the facial expressions (especially tired-looking eyes), pale skin, unkempt physical appearance, and untidy clothes acted as visually perceivable category features that communicated tiredness and exhaustion. Some students associated the features with alcoholism. In this perception, the man was categorized as an 'Irresponsible teacher'.
Judging by the tired-looking eyes and pale skin, this teacher is exhausted. He needs rest and doesn't care what students do during the lesson. He looks messy and is also too tired to take care of his appearance. (Students 58) This teacher is an alcoholic. His untidy appearance, dirty clothes and dark circles around eyes tell about his life-style. He is an irresponsible teacher who isn't interested in students and their learning. (Student 62) For some students, visual features of the paintings generated multimodal experiences. Visually perceivable features became associated with diverse sensory experiences, such as sound and scent.
The lips of this teacher are very tight and somehow the mouth looks angry. He looks strict and silent. He would speak in a weak voice that you can barely hear. (Student 17) His overall appearance is untidy. He smells bad and he smokes. He would be a disinterested teacher who doesn't care about his students. (Student 42) In sum, a number of students perceived the man's facial expression as strict and angry, which generated feelings of fear. Based on these visually perceivable category features (Jayyusi 1984), the man in the image was categorized as a 'Strict teacher'. Other students interpreted the facial expression as tired and the clothes as untidy. For them, these features served as bases of categorizing the 'teacher' as a 'Tired and indifferent teacher'. The membership categorizations 'Strict teacher' and 'Tired and indifferent teacher' generated categorybound expectations that this teacher would keep strict discipline in the classroom or would not be interested in students and would not put effort into the quality of teaching.

Discussion
The findings of this study show that students used visual features of the images of 'teachers' as bases of categorizing them. The visual features acted as 'perceptually available category features' (Jayyusi 1984, 73) that modified the culturally available membership category ('teacher') and contributed to the construction of situational membership categorizations ('certain kind of teacher'). Hence, membership categorization emerged as a process of situational sense-making using cultural membership categories and situationally perceived visual features of the persons as the resources of categorization. This procedure of membership categorization is summarized in The findings of the study indicated that when categorizing the 'teachers' in the images, students used visual features as means of making inferences about the 'teachers.' These inferences related to 'teachers'' personal traits, habits, communicative style, relation to students, and professional competence. In addition to categorizing the 'teachers,' students seemed to simultaneously categorize themselves in relation to them. Thus, categorizing others contributed to selfcategorization (Hogg 2003). It seemed that when students categorized the teachers as certain kinds of teachers with corresponding features and characteristics, they simultaneously constructed for themselves identities as certain kinds of students with similarly category-bound features and characteristics. This notion -provided by the membership categorization analysis in this study -is compatible with relational identity theory (e.g., Schapiro 2010).
The students' emotional responses generated by the 'teachers'' visual expressiveness as well as the inferences they drew about the 'teachers'' teaching and communicative styles implied a certain kind of students' role in the teacher-student interaction. Feelings of fear seemed to restrict their initiative and feelings of trust seemed to encourage them to communicate with the 'teacher.' The 'teacher's' perceived strictness seemed to imply obedience and avoidance of mistakes from the side of students, 'teacher's' perceived friendliness implied uncomplicated communication with the 'teacher,' and 'teacher's' perceived indifference made students assume the 'teacher' might not be interested in students and their learning. Literature on membership categorization analysis has discussed the interactional consequences of categorization stating that categorization influences people's thoughts, feelings and actions related to other people (Bowker and Star 2000;Housley and Fitzgerald 2015;Jayyusi 1984;Lepper 2000), and thus, positions people in relation to each other (Deppermann 2013;Liebscher & Dailey-O´Cain 2013). Based on this study, categorization based on visual cues seemed to have similar positioning consequences.
As this study shows, students categorized the 'teachers' in the images in different ways. Even though the students were all Finnish and, in addition, raised and educated in Finnish schools, they made sense of the 'teachers'' visual cues in diverse ways, which speaks for the occasioned character of cultural categories purported by Francis and Hester (2017). As Francis and Hart (1997) stated, perceivers' backgrounds influence the way they make sense of the visual features and categorize people based on them. A similar notion has been posited, for instance, by dual-processing models of person perception, according to which both visual features and perceivers' knowledge, lived experiences, mood, and motivation shape perception (Brewer 1988;Fiske and Neuberg 1990;Gawronski and Creighton 2013;Macrae and Bodenhausen 2001). Hence, even though culturally shared membership categories may provide a common resource of categorization, the situational variation becomes visible when constructing situational membership categorizations.
Despite the finding that students explicated a number of visually perceivable category features as bases of categorizing the 'teachers' and this study listed them separately in Table 1, it might be relevant to consider that different visual features acted together to form a general overall impression (Adaval, Saluja, and Jiang 2019;Van der Helm 2017). The finding of this study that students also associated non-visual sensory experiences with visual cues seem to speak for this kind of clustering quality of (visual) person perception. However, it seemed that among visually perceivable category features, facial expression, in general, and gaze, in particular, strongly influenced situational membership categorization. This finding of the central role of facial expression in person perception is congruent with prior research (Skowronski and Ambady 2008;Zebrowitz and Montepare 2008).
In real-life situations, people's judgments of each other do not rest only on visually perceivable information, but they have access to multimodal sensory experiences that influence the way of categorization (Adaval, Saluja, and Jiang 2019). In addition, the time-related flow of people's facial expressions, gestures, postures, and movements contribute to the process of categorization. Hence, the situation of categorizing people based on images and live encounters differ from each other. That being said, prior research has recognized the important role people's visual features play in everyday categorization (Fitzgerald 2015;Lepper 2000;Sacks 1972b;Sacks 1992). For example, first impressions -in which categorization is a central operation -often rely on mere visual information (Riniolo et al. 2006). Because such first impressions may have a long-lasting effect on future interaction, either promoting or problematizing it (Gawronski and Creighton 2013;Skowronski and Ambady 2008), people should be aware of the messages they communicate through visual nonverbal behavior and the meanings they construct when perceiving it.
According to Francis and Hester (2017) and Watson (2015), all categories are occasioned in membership categorization analysis. Hence, this must also apply to the culturally available membership category (Jayyusi 1984) 'Teacher' itself. Six membership categorizations were identifiable from the data: 'Friendly teacher', 'Fair teacher', 'Absent-minded teacher', 'Strict teacher', 'Tired teacher', and 'Irresponsible teacher'. These membership categorizations may be used as means of making inferences about how students in this particular study and context understood the category 'Teacher' at a more general level. The findings suggest that the category of 'Teacher' is mainly based on notions of how a teacher relates to students and communicates/ interacts with them. For instance, students did not construct categories such as 'Expert teacher' or 'Knowledgeable teacher', which can also be due to the fact that such categorizations may be more difficult to make based on mere visual features. However, at another level of education, at a university, for instance, the category of 'Teacher' might acquire different contextual meanings and preferences.

Conclusion
Membership categorization is not a neutral activity, but influences people's thoughts, emotions, and actions and shapes people's relations to each other (Lee 2018;Raymond 2019;Ylänne and Nikander 2019). Because it often occurs automatically and unconsciously (Bowker and Star 2000;Eskelinen, Olesen, and Caswell 2010), people might not be aware of the influence categorization can have on social interactions, interpersonal relations, and construction of their own and other people's identities. Because categorizing importantly contributes to the construction of social identities, membership categorization analysis in the context of person perception may provide a deeper understanding of selfhood and otherness. Therefore, we need methods that enable us to identify and critically reflect on our everyday categorizing practices (see Berard 2005;Talmy 2009). The way this study applies membership categorization analysis to person perception provides one possibility to study and raise critical awareness of the categorizing practices we employ when perceiving people.

Disclosure statement
No potential conflict of interest was reported by the author.

Notes on contributor
Jari Martikainen (PhD) is a Postdoctoral Researcher in Social Psychology at the University of Eastern Finland, Finland. His research interests focus on social representations theory, visual and social representations of teachership, visual and arts-based research methods and pedagogy of art history.