M.A thesis. A micro-analytic investigation of gaming interactions: displays of understanding in Overwatch

Abstract

A micro-analytic investigation of gaming interactions: displays of understanding in Overwatch

Full text

Polat, Alper
A micro-analytic investigation of gaming interactions: displays of understanding in Overwatch
Master’s Thesis
FACULTY OF EDUCATION
Master’s Degree Program in Learning & Educational Technology
2020

University of Oulu
Faculty of Education
A micro-analytic investigation of gaming interactions: displays of understanding in
Overwatch (Alper Polat)
Master’s Thesis, 62 pages, 1 appendix
July 2020
This thesis examines the gaming interactions of players in Overwatch, a multiplayer, fast-paced,
first-person shooter video game. The data contains players’ practice matches against other
teams to improve their team synergy, and to try out new additions to their roster. Two hours of
their video-recorded, naturally-occuring gaming interactions in English are analyzed using
Conversation Analysis to uncover insights on how players achieve intersubjectivity (i.e. mutual
understanding). As this phenomenon has not been subject to research in gaming contexts yet,
the analytical focus is on displays of understanding in second turn (i.e. a turn that is addressed
to its prior turn) during both talk and action sequences. Four types of displays of understanding
are identified and discussed in detail: non-understanding, verbal display, embodied display, as
well as embodied and verbal display. The study also discusses players’ engagement in pre-
sequential planning actions, their use of in-game jargon and shorthand expressions. The
findings contribute to the growing body of research on gaming interactions, and to
intersubjectivity research within social interaction. The analyses bring evidence to players
attaining more in-game success when mutual understanding is achieved, and players being more
prone to in-game failures when it is not. Furthermore, discussions regarding players’ speaker
selection and repair practises to overcome interactional troubles are made.
Keywords: displays of understanding, intersubjectivity, multiplayer, gaming, social interaction,
conversation analysis

Contents
1 Introduction ................................................................................................................................................. 4
2 Literature Review ........................................................................................................................................ 7
2.1 Understanding in interaction ..................................................................................................................... 7
2.2 Interaction in multiplayer video games ................................................................................................... 10
3 Methodology ............................................................................................................................................... 14
3.1 Research Method .................................................................................................................................... 14
3.2 Participants ............................................................................................................................................. 16
3.3 Data ......................................................................................................................................................... 16
3.4 Overwatch ............................................................................................................................................... 17
3.5 Transcribing Gaming Interactions .......................................................................................................... 21
4 Data Analysis ............................................................................................................................................. 23
4.1 Displays of Non-Understanding ............................................................................................................. 23
4.2 Successful Displays ................................................................................................................................ 31
4.2.1 Verbal Displays of Understanding ................................................................................................ 32
4.2.2 Embodied Displays of Understanding ........................................................................................... 34
4.2.3 Embodied and Verbal Displays of Understanding ........................................................................ 38
4.3 Summary of main findings ..................................................................................................................... 46
5 Conclusion .................................................................................................................................................. 49
References ............................................................................................................................................................ 54

1 Introduction
For a couple of decades, video games have been a significant source of entertainment for many
people. Game developers have introduced many different types of games (e.g. first-person
shooters, multiplayer online battle arenas, real-time strategy games, sports, and so on). Each
different type has various in-game objectives and they thus provide rich opportunities for social
interaction to take place, as well as problem solving and collaboration skills to be used
(Bluemink & Järvelä, 2011; Gee, 2006). More importantly, as many games are mainly in
English, this means that these skills are manifested in either a second or foreign language by
many players (Sylvén & Sundqvist, 2012). Given this case, the research into gaming contexts
would reveal significant insights on how English is used for problem solving and collaboration
outside of classrooms (Chen & Huang, 2010; Hung, 2007; Sylvén & Sundqvist, 2012). These
insights then can be used in supporting extra-curricular activities in schools, lesson planning,
and curriculum development (Caldwell, Osterweil, Urbano, Tan, & Eberhardt, 2017; Olsson,
2012; 2016; Sundqvist, 2009).
For successful problem solving and collaboration, it is important that group members within
teams achieve mutual understanding (Järvenoja & Järvelä, 2009; Paans, Onan, Molenaar,
Verhoeven, & Segers, 2019; van de Sande & Greeno, 2012). However, group members do not
always reach a mutual understanding easily (van de Sande & Greeno, 2012). This problem gains
a new shape in multiplayer video games that require mobile actions in continuously changing
contexts (Mondada, 2013). In addition, the increased amount of in-game tasks with a time
constraint may make it more challenging to reach mutual understanding. To overcome this
challenge, players of multiplayer video games timely coordinate their in-game actions
(Mondada, 2013; Piirainen-Marsh & Tainio, 2014; Reeves, Laurier, & Brown, 2009). These
timely coordinated in-game actions “not only concern the moment-by-moment unfolding of
actions but also the participants’ anticipations, based on their skilled interpretation in real time
of projected trajectories of actions, both in talk and in the game, as well as the participants’
planning of game strategies within more extensive lines of conduct.” (Mondada, 2013. p. 302).
Some examples to this phenomenon entail actions such as movement, grouping, attacking,
defending, using abilities, and many more, all of which are carried out by player controlled in-
game ‘avatars’ or bodies (Bennerstedt, 2008; Reeves et al., 2009). Other than these bodily
actions, players also use voice chat software to verbally communicate with each other real-time,

5
during which they typically engage in game-related, goal-directed interactions such as
communicating about the objectives of video games (Moore, Ducheneaut, & Nickell, 2007).
These verbal and bodily interactions together constitute significant interactional resources,
through which players achieve ‘intersubjectivity’, a term that refers to mutual understanding
between speakers within a given context (see e.g., Schegloff, 1992).
However, even though these timely coordinated interactions are important for achieving mutual
understanding, they have not been subject to detailed research (Mondada, 2013). To fill this
gap, this study uses Conversation Analysis (CA) as the research methodology. CA is a research
methodology that aims to “describe, analyze, and understand talk as a basic and constitutive
feature of human social life” (Sidnell, 2010b. p.1). Additionally, embodied actions are included
within its framework (Mondada, 2014, 2019; Rauniomaa & Keisanen, 2014). CA provides a
detailed, moment-by-moment analysis that focuses on the sequentiality aspect of social
interaction, which makes it an effective methodology to inspect how mutual understanding is
achieved in social interaction (Schegloff, 1992; Sidnell, 2010). To these ends, this thesis will
aim to shed light on the timely coordinated interactions through which players attempt to arrive
at mutual understanding as they play a multiplayer first-person shooter game called Overwatch.
Detailed observations will be made of players' verbal interactions intertwined with their in-
game avatars’ bodily interactions. This will be done because in the game world players are
represented by their in-game avatars and they orient themselves to their avatars’ bodily
interactions as if they are real bodies.
In order to understand how players arrive at mutual understanding in gaming contexts, the
present study focuses on ‘displays of understanding’ in the second turn (i.e. a turn that is
addressing a prior turn) within in-game talk and action sequences (e.g. talk-centered planning
situations, fast-paced combat situations). Furthermore, the cases in which understanding is not
displayed are also discussed. Displays of understanding refer to ‘relevant’ verbal turns and/or
embodied actions deployed by participants to demonstrate that they have understood what is
going on in the interaction (see e.g. Sidnell, 2010. pp. 59-76). More detailed explanations of
this phenomenon will be provided in the next chapter alongside the gaming behavior of players.
The research questions guiding the analyses are:
1- What types of display of understanding do participants use in Overwatch?

6
2- What are the interactional resources used by the participants during display of
understanding moments?
3- How does displaying understanding affect players’ in-game success?
In the third chapter, the research methodology, its methodological principles, as well as the
data, and transcription procedures will be further elaborated on. After that, in chapter 4 the
analyses and findings will be presented. Lastly, in chapter 5, the importance of these findings,
limitations of the present study, and suggestions for future research will be discussed.

7
2 Literature Review
In this chapter, I will set the grounding of this thesis based on discussions of the literature in
two sub-chapters. With the first sub-chapter, I will discuss the CA work on understanding in
interaction. In the second sub-chapter I will discuss research on gaming interactions.
2.1 Understanding in interaction
In CA, “understanding” can be understood by looking at the reciprocal relationship between
two turns-at-talk. For instance, when the person A asks “Would you like to drink something?”
a potential answer by the second person B can be a range of things like accepting/declining
“Yes/no”, or specifying what B would like to drink “Water, soda” etc. instead of what might
sound odd to A such as “I am fine, how about you”, which could be more relevant if A was
asking “How are you?”. So, here it is seen that in the first turn that A produced, s/he has made
relevant a set of responses for B to choose from, and respond. The second speaker B, while
responding to A’s question with a relevant response, will, therefore, show his/her own
understanding of A’s turn simultaneously. Furthermore, if in the second turn there is an
unexpected response, A might extend the talk by another turn and try to fix the issue by doing
a range of things such as rephrasing, clarifying, asking another question, and so on. In other
words, understanding is concerned with “... what a second participant does in response to a first
participant” (Mondada, 2011. p. 543). Participants, in the second turn address the first,
continuing with actions that show their understanding (Heritage, 2009). Consequently, actions
can be interpreted as what participants do with their turns-at-talk (Drew, 2013; Sidnell, 2010).
For instance, when someone asks “are you thirsty?” the purpose of the turn could be merely an
inquiry, or to offer something to drink, when the ensuing interaction is observed the purpose
can be uncovered.
Within CA, this aforementioned relation of two turns-at-talk is referred to as “adjacency pairs”
(Schegloff & Sacks, 1973). An adjacency pair consists of two pair parts: a first pair part, and a
second pair part. Examples of such paired actions are questions-answers, greetings-greetings,
offer-accept/decline, and alike (Schegloff, 2007). When a first pair part has been produced, it
makes the second pair part relevant, as Schegloff (1968) notes: “given the first, the second is
expectable; upon its occurrence it can be seen to be a second item to the first; upon its
nonoccurrence it can be seen to be officially absent - all this provided by the occurrence of the

8
first item.” (p. 1083). For example, these are actions that can be questions-answers, greetings-
greetings, offer-accept/decline, and alike (Schegloff, 2007). After the second pair part has been
completed, the first speaker can do one of two things: accept the response and continue, or point
to a trouble source. This occurs in the next turn after the second pair part, which is called “third
position” (Heritage, 1984). If a response in the second pair part is not in line with the first pair
part, it will likely be treated as a problem source by the first speaker and be subject to a repair
attempt (Schegloff, Jefferson & Sacks, 1977; Schegloff, 1992) in the third position (Heritage,
1984) so that understanding can be restored. Repairs refer to interactional mechanisms deployed
by participants to overcome interactional problems that may stem from speaking, hearing, and
understanding (Hayashi, Raymond & Sidnell, 2013). For instance, repair can be in the form of
asking a clarification question, asking for repetition and so on. Furthermore, participants hold
each other accountable for putting the relevant actions turn-by-turn, which is called sanctioning.
As Sidnell (2010) asserts: “Adjacency pairs allow then for a framework of understanding that
is constructed and sustained on a turn-by-turn basis” (p. 66).
In social interaction, speakers carry out actions not just by speaking, while they speak they
simultaneously use various resources of the body (Goodwin, 2000; Hindmarsh, Reynolds, &
Dunne, 2011; Lindwall, 2014), such as gestures (Mori & Hayashi, 2006; Belhiah, 2012), and
gaze (Korkiakangas & Rae, 2014). Prior studies have shown how embodied resources are
consequential for achieving understanding in interaction (Goodwin, 2000; Hindmarsh, et al.,
2011; Lindwall, 2014; Mondada, 2014b; Mondada, 2019). For instance, a second pair part in
an adjacency pair may include embodied actions to demonstrate that the second speaker has
understood the first pair part, and provided a relevant response (e.g. nodding, see Stivers, 2008;
2010; Stivers, Mondada & Steensig, 2011 for more instances). Furthermore, a second pair part
may be entirely made up of embodied actions (Rauniomaa & Keisanen, 2012). Therefore, this
‘adjacency’ should not be necessarily limited to turns-at-talk, but include embodiment as well
(Lindwall & Ekström, 2012).
Building on these earlier research, we can claim that the understandings shown in the second
pair part can have characteristics such as (1) verbal, (2) embodied, (3) both verbal and
embodied. Furthermore, there can be (4) non-understandings or misunderstandings that can lead
to troubles in interaction (e.g. delays, silences, etc.). More specifically, non-understandings are
characterized by a missing second pair part or a relevant second action whereas, in
misunderstandings the second pair part exists, but it does not provide the expected response to

9
the first pair part due to reasons such as misinterpretation of the first pair part (see Schegloff,
1987 for more examples).
In sum, we know that understandings are made visible in the second pair part using a variety of
interactional resources, and are treated as sufficient or insufficient (e.g accepted, declined,
repaired, etc.) in the third position. But how is understanding displayed? While displaying their
understanding, the speakers tend to either merely claim that they have understood (e.g. “oh-
prefaced turns” see Heritage, 1984), or demonstrate their understanding (e.g. make an
interpretation). Consider the following example from Sacks for more clarification on this
matter:
(Sacks, 1992:II:141)
1 A: where are you staying
2 B: Pacific Palisades
3a A: oh at the west side of town
vs
3b A: oh Pacific Palisades
We see that in 3a, A makes an interpretation of what B said in line 2 by re-phrasing the location.
By doing so, A demonstrates understanding and this enables B and the analyst to infer that A
knows where Pacific Palisades is. Whereas in 3b, A simply claims understanding by merely
repeating the said location. Hence, from 3b we cannot make analytical claims that A knows
where the said place is located. I will establish my observations based on this explanation of
displaying understanding. My analysis will be concerned with demonstrations of understanding
as “displays of understanding” rather than claims, for claims arguably do not provide a fully
relevant action in the second pair part when in-game actions are considered.
In normal conversation, speakers treat understandings as taken-for-granted, which means that
they do not explicitly orient to understandings unless there is a problem with them (e.g. a person
would not ask “do you understand?” unless they pick up a set of clues that point to problems in
the second speaker’s turn). In other words, it can be said that even though understandings are
almost always there in the conversation, they have an invisible status (Schegloff, 1992;
Seedhouse, 2004; Mondada, 2011). However, the analyst can access and analyze understanding
by looking at the sequences in which turns and actions are produced (Seedhouse, 2004;

10
Mondada, 2011), or by looking at the preceding turn that made the second turn relevant (Sidnell,
2010). In the analysis section, I will be documenting the gaming interactions in a similar way.
To summarize, I have reviewed how “understanding” is perceived and studied in CA research.
The emphasis has been on adjacency pairs framework. Within this framework, understanding
is displayed in the second pair part to a relevant first pair part (e.g. question, request, summons,
directive, instruction, etc.) by using different interactional resources (e.g. verbal, embodied).
Moreover, I have shown the difference between claims and demonstrations of understanding,
to which I will further refer when considering displays of understanding in my analysis. Lastly,
I briefly elaborated on the status of understanding in conversation (i.e. taken-for-granted), and
explained how my analysis is going to treat this phenomenon while exploring understanding in
gaming interactions. Next, I review the literature on gaming interactions and point out the
research gap I will try to address.
2.2 Interaction in multiplayer video games
This section will address studies of interaction in multiplayer video games. That said, first it is
relevant to first consider what players generally do in games, then consider particular properties
of physical location of players, for it will have an impact on how they interact with each other.
I will start by noting the players’ general in-game behavior regardless of setting. Then, I will
move on to describe the setting-specific affordances for interactions (i.e. what the players can
do while interacting). For that purpose, I will discuss co-present as well as geographically
dispersed settings. At the end of this chapter, I will provide a brief summary of this chapter and
point out the research gap.
In multiplayer video games, players come together in virtual worlds (e.g. a map or game world)
(Reeves, et al., 2009; Berger, Jucker, & Locher, 2016) to constantly carry out coordinated
actions such as “grouping”, “fighting”, “moving”, and “waiting” (Bennerstedt & Ivarsson,
2010), and many more to fulfill in-game tasks such as defeating an enemy team or a group of
in-game monsters. While doing so, each player takes on different roles (e.g. tank, damage,
healer) (Bennerstedt & Linderoth, 2009), and attempt to competently execute their role-specific
actions to assist their teammates (Ducheneaut & Moore, 2005). During their coordination, the
players refer to each other in various ways. Players either address (1) the player directly, (2)
their in-game avatar, (3) the player and their avatar together (Mondada, 2013; Baldauf-

11
Quilliatre & Colón de Carvajal, 2015). They use in-game jargon to guide their in-game actions
more concretely and smoothly (Wright, Boria & Bradenbach, 2002; Bennerstedt, 2008). Lastly,
after long hours of playing with each other, the players are shown to develop an “interactional
synchrony”, a rhythm that allows them to coordinate their actions timely and effectively
(Piirainen-Marsh & Tainio, 2014).
When players are co-present with each other, they tend to play games on either the same screen
(see; Aarsand & Aronsson, 2009; Piirainen-Marsh & Tainio, 2009; Mondada, 2011b, Mondada;
2013), or they set up multiple screens in the same room (see Sjöblom, 2008; Keating &
Sunakawa, 2010). This means that the way they interact with each other does not get limited to
the on-screen, in-game actions. In other words, they still use their real bodies (i.e. embodied
actions such as gestures, gaze, etc.) and talk to each other (i.e. verbal actions) when interacting
with each other. While playing, players use a range of verbal resources such as prosody, deixis
(e.g. pronouns), repetition, as well as embodied resources such as gestures, gaze, pointing
(Keating & Sunakawa, 2010). These interactional resources are used in activities such as
questioning, analyzing, observing, or commenting on the co-present players’ gameplay
(Sjöblom, 2008; Tekin & Reeves, 2017). When the game is suspended, players retract to rest
or do other real-life activities (Mondada, 2011b; Mondada, 2013).
To summarize, we can infer that when co-present, players are able to share the same screen or
have access to their co-present players’ screens. Hence, they can use their real-life bodies more
effectively while interacting with each other, which can eliminate many obstacles in
communication. For instance, while referring to something on-screen, they can point at it to
clarify any misunderstandings, or they can make sure that their co-present player is looking at
the same thing/same direction as they are, for “looking” can be significant for in-game
collaboration (see Reeves et al., 2009). Moreover, because they can talk to each other in real-
time and convey their message coupled with these embodied actions, we can claim that the
players have more interactional resources to work with than geographically dispersed settings.
While players are geographically dispersed, they do not have access to each others’ screens
physically. Yet, they can still see each others’ avatars and where these avatars are looking, or
which embodied actions these avatars are carrying out as far as the game in question allows
(see Manninen & Kujanpää, 2005; Bennerstedt, 2008; Bennerstedt & Linderoth, 2009). For
instance, a player can interact with other players via their in-game avatars’ embodied actions

12
(e.g. waving, greeting, etc.). The player can also infer the direction to which another player is
looking from their in-game avatars, or whether their in-game avatar is fighting, resting, dancing,
jumping, or doing any other embodied action that the game supports. Moreover, for games that
do not necessitate fast-paced gameplay, players tend to use in-game text chat (see Bennerstedt,
2007 for an example), and for the fast-paced games, they prefer using built-in voice chats that
are in the game or use other voice chat software (e.g. TeamSpeak or Discord) (Moore, et al.,
2007).
In other words, while geographically dispersed, for verbal interactions players rely on voice
chat software, and for embodied actions players rely on the in-game avatars. Even though the
voice chat can be a sufficient substitute for verbal interactions, in-game avatars are still far from
accurately and fully representing real-life embodied interactions, or conveying the actual state
of their controlling players (Kohonen-Aho & Vatanen, 2020; Manninen & Kujanpää, 2005;
Moore et al., 2007). Furthermore, the players do not have access to each others’ screens, which
creates an information gap (see Balaman, 2015). In other words, a player does not have access
to see what another player sees on their screen, this information can be, for instance, that
player’s abilities (e.g. their ability cooldowns - the time required for an ability to recharge after
being used). Limitations such as these will have a significant impact on the interaction between
players, even more so if the games are fast-paced. This is because that the players will have to
communicate with each other many things (e.g. abilities, enemy locations, game plan, etc.)
simultaneously as they are trying to complete the in-game objectives (see Reeves et al., 2009,
Mondada, 2011b, Mondada, 2013).
To summarize, gaming research successfully shows the following: players engage in a diverse
set of in and out of game interactions while playing games, they make use of game jargons,
they value in-game competence, they attempt to collaborate with each other to fulfill in-game
tasks, and they do not limit their communication to in-game text chat, but use other software to
talk and interact with each other. However, detailed studies remain scarce on how players timely
coordinate their activities in gaming (see also Mondada, 2013). As Reeves, Greiffenhagen &
Laurier (2017) state, the in-game actions “are sequentially organized and environmentally
positioned by players in a purposeful, concerted way” (p. 22). This paves the way for a detailed
analysis of these interactions. Therefore, the scope of this study is to examine these “timely
coordinated” (Mondada, 2013), sequentially organized and environmentally positioned
purposeful in-game actions. To my knowledge, there is currently no research addressing

13
players’ displays of understanding from this perspective. So, I will focus on how players display
their understanding of each others’ actions, which is key to establishing mutual understanding
between themselves. I also aim to contribute to the literature with explorations of voice chat as
a medium of talk, as most of the previous research has focused on in-game text chat in
geographically dispersed settings and real-life talk in co-present settings.

14
3 Methodology
In this section, I will elaborate on the research method used for the study by pointing out its
basic principles, as well as analytical procedures. I will also provide detailed information about
participants and data. I will explain what the data contains and how it was transcribed with
minute detail. Lastly, I will talk about the transcription conventions before transitioning to the
analysis.
3.1 Research Method
In this thesis, Conversation Analysis was used as the research method. It is a process-oriented,
data-driven, qualitative research methodology that studies naturally occurring interactions with
the aim of uncovering patterns through which people interact and behave (Sidnell, 2010b), in
contexts ranging from everyday life to institutional (ten Have, 2007). To uncover these patterns
of interaction, the analyst’s duty is to investigate how participants enact interactional practises.
These interactional practises are: (1) turn-taking, (2) sequencing, (3) overall structuring, (4)
repair (Wong & Waring, 2010). Turn-taking is the fundamental practise upon which
conversation is built, it is concerned with how participants construct and allocate turns (Sacks,
Schegloff & Jefferson, 1974).
Sequencing practises are more general practises than turn-taking, they are ways through which
participants initiate talk, or respond to talk, while carrying out actions like greetings, story-
telling, making requests, invitations (Sidnell, 2010; Wong & Waring, 2010). Overall structuring
practises are concerned with how participants open or close sequences of talks (Sacks, 1995).
For instance, saying “hi” to initiate a greeting sequence, or “don’t I know you from
somewhere?” as a part of so-called “pick-up lines” (see Sacks, 1995). Repair practises refer to
actions enacted for solving interactional problems or “infelicities” (p. 1) that may stem from
speaking, hearing, understanding (Hayashi, Raymond & Sidnell, 2013). Repairs are initiated to
point out a problem source, they are finalized when there is a “solution or abandonment of the
problem” (Schegloff, 2000. p. 207).
Conversation Analysis employs an emic approach to the analysis, in other words participant’s
perspective. Therefore, the analyst’s aim is to uncover the aforementioned interactional patterns
strictly from the talk-in-interaction (i.e. the data itself) rather than consulting external theories

15
or explanations (Seedhouse, 2005). To gain access to participant’s perspective, it is significant
to consider the adjacency pairs, context of the talk, and the sequentiality. Observing adjacency
pairs allows the analyst to uncover the participant’s interpretation of a prior turn within the
context it was made available to all of the participants of that interaction. Observing
sequentiality grants the analyst participant’s perspective on what the participant deemed
relevant in any point of the interaction. Hence, all of the claims that are made from the data are
strictly based on the participants own interpretations of the interaction and their contributions
to it rather than a pre-developed set of coding categories or external theories. Furthermore, the
data is transcribed in minute detail in order to accommodate the need to gain access to
participant’s perspective, as well as to make the data available for scrutiny for readers
(Seedhouse, 2004).
At the beginning of a conversation analytic research, the analyst makes an ‘unmotivated
looking’ into the data to catch any phenomenon of interest (ten Have, 1997; ten Have 2007). In
other words, the analyst approaches the data without any prior thoughts with regards to what
might come out from the data, so that no potentially interesting phenomenon is missed. After
something of interest has been found, the analyst starts looking for other similar instances of
that particular interactional practice so that a sufficient collection of instances can be made.
While making these collections, the analyst can see how that interactional practice is enacted
by the participants. Consequently, the analyst is able to describe in detail the enactment process
of this interactional practice within that particular context as well as out of that context. More
specifically, as Sidnell (2013) notes, the analyst can describe “generic, context-independent
properties” of the interactional practice (p. 78).
In my thesis I carried out the analysis in following steps:
1- Unmotivated looking
2- Noting possible interesting phenomenon for research
3- Making basic transcriptions
4- Choosing an interesting phenomenon for research
5- Making a collection
6- Making detailed transcriptions for the collections
7- Making detailed analyses of each example

16
After I collected the data, I watched the entire data repeatedly to notice an interesting
phenomenon to study. My initial personal interest was to try to find a phenomenon relating to
the development of second language interactional competence. It is important to note that
unmotivated looking refers to a certain kind of “open-minded” approach to the data. In other
words, even though the researcher has some broad ideas or interests in mind, they do not only
focus on finding a phenomenon related to those. Hence, my initial notes of interesting
phenomenon were mostly related to this framework, but were not limited to it. Next, I made
basic transcriptions. Then, after formulating more ideas I finally came to the conclusion that I
should research how players understand each other during the fast-paced, chaotic in-game
moments instead of second language interactional competence matters as the data seemed not
suitable for it. So I started focusing on ‘displays of understanding’ and started making
collections where these took place. To find a diverse set of examples to capture this basic
phenomenon of interaction, I focused on the most basic and salient features of displays of
understanding. These formed the categories that I included in my analysis section. Next, I made
detailed transcriptions of the instances that I collected within each category so that I could make
minute-detail, sequential, robust analyses. Making detailed transcripts also aims at making the
data available to other researchers for more scrutiny and transparency. Lastly, I made detailed
analyses of every example within all categories.
3.2 Participants
There are eight participants from various European countries (i.e. United Kingdom, France,
Germany, Norway, Slovakia). All of the players play the game in English. Except three of them,
all are non native speakers of English. Their ages range from 15-27. Every participant except
the oldest one is studying in high school at the time of data collection. All of the participants
are competitive players with high ratings in Overwatch.
3.3 Data
As the nature of this research suggests, the collected data is in digital video format. Recordings
of multiple online gaming sessions were acquired from one of the participants (TIM) over two
weeks. These reached a total of 14 hours video data, 2 hours out of which were transcribed and
analyzed. The participants gathered up online multiple times in two weeks during which they
played Overwatch as a team and practiced against other teams, which is called “scrimming” in

17
the gaming society. In other words, two teams arrange matches during which they try to utilize
various tactics against each other to improve their understanding of the game as well as their
team synergy. Therefore, the matches are only for practising purposes and have no rewards or
punishments at the stake. Scrimming is also done in order to find new members to a team that
is recruiting new players to their ranks. This means that in some of the matches one or two
participants were swapped out with other participants. All the recordings are from the
perspective of TIM, who is the support player of his team.
3.4 Overwatch
Overwatch is a multiplayer, team-based, first-player shooter (FPS) game. The game is centered
on weapon-based combat in a first-person perspective, which means that the players experience
the game take action from the eyes of the protagonist/hero. These heroes are in-game avatars
that represent the geographically dispersed players. Each hero has a unique name (e.g. Mei,
Reaper, Lucio) and its own unique set of abilities (e.g. Ice Wall, Death Blossom, Speed Boost)
to be used in the battle. There are three main roles in the game: tank, damage, and support.
Tank role mainly focuses on employing the front lines in battle, to take damage and attention
from enemies, create space for their own team to deal damage and kill the enemy heroes to
secure the objective, while damage role as its name suggests, focuses on dealing damage and
removing enemies from the battle, and lastly support role makes use of various skills to keep
their teammates alive, assist them to be more efficient in getting rid of their enemies. Each of
the aforementioned roles is crucial for a team’s ultimate goal: to secure the victory. Hence, the
game’s audience is those players who wish to compete with others on various settings through
diverse game mechanics and according to a set of collaborative objectives (see below for a
detailed discussion of objectives). Action based nature of Overwatch obliges players to
communicate with their teammates quickly real-time, make prompt decisions to overcome their
enemies and carry out the aforementioned missions various game modes contain. To carry out
these missions, players tightly coordinate their actions with other players while they are
grouping, fighting, defending, hiding and such. Moreover, there is a built-in voice chat feature
in the game that allows players to speak with each other. Hence, the players develop novel ways
of communicating through the uses of “shorthand” or “rapid” expressions. For instance, to talk
about the low health status, players may say “he is one” meaning that “he is at one health point”,
which indicates that the enemy health point is dramatically low and that particular enemy can

18
be removed from the battlefield without hassle. These expressions will be explicated further in
the analysis section.
As mentioned earlier, there are various in-game objectives depending on the chosen game
mode. These game modes are: assault, control, escort, hybrid. Each game mode can be played
on its own unique set of in-game locations (i.e. ‘maps’) (Reeves, et al., 2009). Moreover, game
modes are named after the in-game objectives they contain, hence the names for game modes
are interchangeably used with the in-game objectives. The game modes are further elaborated
below:
1. Assault: In assault, the attacking team must make two capture points. The defending
team tries to protect the points and prevent the attackers from making these capture
points until the attacking team’s time runs out. When a point has been captured by the
attacking team, their time gets extended. At the end of a round, teams get swapped and
the attackers start defending, while defenders start attacking.
2. Control: In control, both of the teams are attempting to capture a point and hold the
control of it until they reach 100% completion. When the point is captured by a team,
they start defending it from the opposing team so that they can make it to full completion
before their enemies. Control game modes are in best-of-three format, in order to win,
a team must make two full completions. Unlike the other game modes, this mode has
no time limit.
3. Escort: In escort, the attacking team has to deliver the payload to a final checkpoint. The
defending team attempts to prevent this by blocking the path of the payload. Each escort
map has two to three checkpoints that attackers must make through, the time gets
extended once a checkpoint has been reached by the attackers. The team that delivers
the payload further than the other gets the victory.
4. Hybrid: Hybrid modes are a mixture between assault and escort. The attacker team must
capture a point, then deliver the payload to a final checkpoint while defenders try to
prevent this from happening.

19
In-game interface involves multiple features that players constantly orient to while playing.
These are described with the picture below:
1. Hero’s status bar - It contains a small picture of the hero, their health points (200/200 in
this case), and their on-fire bar, which indicates how well the player is performing at
that particular moment.
2. Hero’s ultimate ability status - It shows the percentage of power accumulated so that
the hero’s unique ultimate ability can be used. Players accumulate this power by dealing
damage to the enemy team, or healing their own teammates. Ultimate abilities can only
be used when this circle fills up and 100% has been reached. Right below the circle, the
key assigned for the ultimate ability is shown.
3. Hero’s normal abilities & ammo - Besides ultimate abilities, each hero has its own
unique set of normal abilities. Each ability is displayed with a small picture, with
assigned keys beneath them. The abilities need to re-charge for a while once they are
used. While they are re-charging the picture shows a number that represents the time (in
seconds) required for the ability to be available. On the right side of the abilities, the
hero’s ammo is shown. Heroes can attack their enemies with a limited amount of ammo
(15/15 in this case) before they have to reload. Some heroes have unlimited ammo and
do not need to reload, if that is the case, it is represented with an infinity symbol (∞).

20
4. Allied and Enemy heroes - The avatars of allied and enemy heroes are shown. Above
an avatar, it’s health status is displayed with a small white bar. Below it, there are the
nicknames of players (censored with blue for allies, red for enemies). For players to
distinguish an ally from an enemy, enemy heroes have a small red lining surrounding
them.
5. Player’s crosshair - This small green circle indicates where the player aims with their
weapon. In FPS games, the crosshair is always fixated right in the center of the screen.
6. Objective status - Here, the objectives are shown as well as the time remaining for the
completion of them.
7. Kill feed - Information related to which hero has been killed by whom, using which
abilities are shown here. Each player in the game is notified once someone dies, or
when their ultimate ability does something (e.g. killing another hero, getting
destroyed).
8. Objective location - The in-game location of the objectives are shown with letters
representing them. Additionally, borders of the locations are highlighted with colored
lines around them (blue for ally controlled, red for enemy controlled). For escort maps,
the direction that a payload follows is highlighted.
Overwatch was chosen as the context to analyze interaction for multiple reasons. First of all,
the game design creates room for interaction by requiring players to take on and fulfill different
roles, which promotes interaction to occur. This interaction, however, is elusive to examine due
to the fast-paced nature of the game. As they try to get rid of their enemies, players have to keep
in mind many things such as their own abilities, their teammates’ abilities, enemy abilities, their
locations, and many more. Hence, players find themselves in a chaotic in-game battle
environment where they need to carry out their role specific tasks while simultaneously
communicating the information they have with their teammates and achieve a synchrony.
Therefore, for a team to be successful, each player must timely coordinate their actions with
their teammates (Mondada, 2013). Because the research on these timely coordinated
interactions is scarce, the present study aims to contribute to fill that gap. Finally, the current
popularity of Overwatch made it a relevant source to examine. Even though the game’s
publisher Blizzard has not revealed the actual number of players in 2019, it was announced to
be around 40 million back in 2018. In the next section I will explain how the transcriptions were
made, and illustrate the details they contain.

21
3.5 Transcribing Gaming Interactions
Conversation Analysis as a methodology requires the transcription of naturally occurring video
data for detailed moment-by-moment analysis. For the data to be presented as complete and
accurate as possible, two transcription conventions were used together. Namely, Jefferson
(2004) method for transcription of talk, Mondada’s (2014) approach for transcription of
embodied actions. Both of these transcription conventions were developed for face to face
interaction, however, in the current study they are adapted to transcribe the talk of the players
and embodied actions of video game avatars which are designed to be somewhat representative
of real life interactions (Berger et al., 2016). See Appendix 1 for the full list of markers and
symbols used throughout the data.
Transcript example below will be used to illustrate the transcripts presented for analysis in this
thesis. Even though their in-game nicknames serve as a means to anonymize them, pseudonyms
(i.e. ATI, DEN and so on) will be used in all transcripts in order to protect the privacy of the
participants, as well as to ease the reading process. Before the numbered lines, all the heroes
that exist during that particular transcript, map and the game mode are noted. Even though each
team consists of six heroes, not all of the heroes are listed to preserve clarity and readability.
The numbered lines present the talk and pauses as they occur in the data. The first three letters
after a line number indicate the speaker. If there is an embodied action taking place within a
turn, those will be marked using various symbols (i.e. +, %, &, etc.), and the corresponding
embodied action will be demonstrated below the spoken turn. According to the Mondada
method (2014), an embodied action begins where a symbol is placed in the corresponding turn
(see line 1). The duration of the embodied action is represented by a continuation arrow (--->),
until the same symbol is reached, which indicates that the embodied action has ended (see line
3). Figures indicate the screenshots taken at the exact moment in the turn, marked with square
(#). All of the lines where embodied actions are indicated begin with the speaker’s pseudonym,
in order to identify who the action was carried by. Lines where figures are attached begin with
a short form of figure (fig), and the figures are numbered.
Consequently, embodied actions indicate the actions of the in-game avatars rather than the
players themselves. These actions involve a variety of in-game actions such as gaze (i.e. where
the in-game avatars look), ability usage (i.e. special abilities of these avatars), movement (i.e.
avatar movements), and so on. To preserve the readability, movements and gaze were only

22
transcribed wherever it affected the analysis. Moreover, since the current study only had access
to TIM’s point of view in the games transcribed, the access to other players’ gazes and
embodied actions is limited.
Excerpt0: Example transcript
Map: King’s Row (Hybrid)
Heroes: TIM (Lucio), DEN (Orisa), ATI (Mei) …
Enemy heroes: None
1 ATI: oh they’re #+%up there up there
ati +shoots upwards-->
den %shoots upwards-->
fig #fig.1
2 (0.2)
figure 1
3 DEN: +%*they’re coming left *they’re coming left*
ati -->+
den -->%
tim *looks upwards-------*looks left---------*

23
4 Data Analysis
This section will illustrate the analysis carried out for this thesis. As emerged from the
data, various types of display of understanding were identified. These were first grouped into
two main categories: Displays of Non-Understandings, and Successful Displays. There were
two reasons for this choice. First, while the first category encompasses cases of interactional
troubles and how these are dealt with, Successful Displays focus on cases where mutual
understanding is achieved by players. Second, these main categories have different interactional
and game-related consequences as will be emphasized in the analyses. Then, the second
category was further seperated into three sub-categories. These are namely: Verbal Displays of
Understanding, Embodied Displays of Understanding, Embodied and Verbal Displays of
Understanding. In sum, display of non-understanding indicates that there has been a
demonstration of non-understanding either by a mere claim of understanding rather than a
demonstration of it, or by the entire absence of a second pair part. This second pair part refers
to potentially relevant actions either with verbal or embodied resources. Second, verbal display
indicates that the understanding has been demonstrated during a verbal turn. Third, embodied
display indicates that the understanding has been demonstrated using an embodied resource by
the player controlled avatars (i.e. gaze, ability use, movement, etc.). Lastly, embodied and
verbal display indicates that the understanding has been demonstrated using both verbal and
embodied resources simultaneously, or in contingence (i.e. in close proximity time-wise).
The analysis will be presented in the following order: (1) introducing the background for the
excerpt, (2) presenting the excerpt, (3) turn-by-turn analysis of the excerpt, (4) the analytical
significance of the excerpt. Within the same category, the excerpts presented after the first one
will be used to describe the variance in that category. In other words, the excerpts within the
same category will differ from each other in certain aspects, these will be elaborated on while
introducing the background for the following excerpts after the first one was presented.
4.1 Displays of Non-Understanding
This section focuses on excerpts where interactional troubles occur and lead players to not
achieve mutual understanding, making them more prone to in-game failures. Both non-
understanding and misunderstanding moments are analyzed. Particular attention is paid to the
events leading up to the potential cause of this phenomenon (overlaps, no-hearing etc.), as well

24
as to the events following the display of non-understanding moments (i.e. sanctioning, repair).
Players usually ignore these moments arguably because there are no negative consequences for
the team, or when there are other immediate focuses in-game. However, sometimes, when these
hurdles occur, the in-game consequences for the team are grim. For instance, a teammate might
die because of it, or an attack/defense might fail. Furthermore, these moments may trigger
upcoming interactional sequences during which remedies are occasionally attempted to solve
the non-understanding case.
In the first excerpt, I show a case where a display of non-understanding gets ignored. The team
is escorting the payload to the second checkpoint, during which they encounter enemy pharah
shooting from above.
Excerpt1: Save ults
Map: King’s Row (Hybrid)
Heroes: ALP (Sombra), MIT (Lucio), KLD (Orisa)
Enemy heroes: Pharah
1 DEN: we might lose so lets save ults.
2 ELO: yeah don’t barrier [here.]
3 DEN: [yeah ] △dont.
tim △uses ultimate
4 DEN: oh why-
5 ATI: >we can win ⌂we can win.<
col ⌂kills enemy doomfist
6 COL: >i killed doom< keep going.
7 TIM: yeah i didn’t hear.
8 DEN: [no problem.]
9 COL: [it’s ok we ] won.
In line 1, DEN suggests to save ‘ults’, which refers to saving the remaining ultimate abilities of
his teammates for an upcoming teamfight. For this, he gives the reason that they might lose the
ongoing teamfight. ELO in line 2 follows this decision by saying “yeah don't barrier”, which is
TIM’s ultimate ability. As he is finalizing his utterance, he overlaps with DEN saying “yeah”,

25
as DEN continues with his turn, TIM uses his ultimate. DEN in line 4 questions this decision
by saying “oh why-”, ATI cuts him off by quickly saying that they can win the teamfight
deemed as potentially lost by DEN earlier. During ATI’s quick utterance, the team is informed
of COL removing enemy doomfist from the teamfight. COL also utters this fact in line 6, and
reinforces his team’s morale by telling them to “keep going”. In line 7, TIM gives an account
of why he used his ultimate “yeah i didn’t hear”. This is the first time in this interaction where
an explanation is made regarding not having listened to an earlier callout. Given TIM’s
explanation, we can claim that TIM displayed a non-understanding with his ultimate usage. In
other words, TIM has not demonstrated the relevant action of keeping his ultimate for the next
teamfight, and he gives an account of this. However, the team seems to not treat this non-
understanding as a negative case. DEN and COL give a ‘by-pass’ to this non-understanding by
saying “no problem”, and “it’s ok we won”. Here, players arguably ignore the display of non-
understanding by TIM and they do this because the in-game consequences for the team were
not grim. In other words, despite the earlier callout by DEN, the team has won the teamfight.
This excerpt is significant in demonstrating how players ignore some of the non-understanding
moments. The main reason for this seems to be the players’ assessment of in-game
consequences. DEN and COL assess the situation as non-problematic, and because it is deemed
as not negatively consequential for the team, the non-understanding is ignored.
The following excerpt will demonstrate an example in which a display of non-understanding
by one player (ELO) takes place, but this time it is not ignored. The fact that non-understanding
has been displayed is going to be sanctioned by another team member.
Here, the team is trying to push the payload to the second checkpoint. They are almost there
but there is one final stretch, at which their enemies are attempting to stop them.
Excerpt2: Said that
Map: Blizzard World (Hybrid)
Heroes: TIM (Lucio), ELO (Moira), SIN (Sigma), DEN (Orisa)
Enemy heroes: Reaper, Orisa
1 TIM: #they are stacking high ground.
fig #fig.2

26
2 *#(0.2) *#(1.0)*#
tim *turns behind*turns in front then quickly turns behind*
fig #fig.3 #fig.4 #fig.5
figure 2 (looking in front) figure 3 (looking behind)
figure 4 (looking in front) figure 5 (looking behind)
3 TIM: >REAPER BEHIND [+↑TP- R- &BEHIND TP BEHIND↑<]
4 DEN: [+>yeah &i’m bongoing i’m
5 bongoing.<]
tim +shoots at reaper--->
den &uses ultimate
6 (1.5)
7 DEN: [>orisa one*+ orisa one.<]
8 SIN: [orisa orisa.]
9 (0.5)
tim *turns to orisa
tim -->+shoots at orisa--->
10 ●(0.2)

27
elo ●dies to reaper
11 TIM: +△ [beating].
12 ELO: +△ [oh reap]er behind.
tim -->+△uses ultimate
13 (2.0)
14 DEN: yeah TIM said that.
In line 1, TIM signals to his team that the enemies are gathering up on the high ground. On his
screen, we see that he is referring to the balcony above his team (#fig.2). At the end of his turn,
he looks behind to assure there is nobody (#fig.3). After this, he looks in front of himself to
check his enemies (#fig.4), then looks behind himself again. As he looks back, he spots the
enemy Reaper teleporting behind his team. Reaper’s teleportation looks similar to a red shadow
as can be seen (#fig.5). In line 3, TIM produces a first pair part with his warning action; he
shouts and warns his team that one enemy player is right behind them, effectively making
relevant a second action (i.e. acting carefully according to the warning). Players use warnings
like these so that their entire team knows the danger in their backlines, and reacts to it
appropriately either by taking the enemy out, or by running away from that enemy. The physical
location of players’ in-game avatars make it even more relevant for them to do something about
a threat if they are closer to that threat (e.g. Reaper). Other players except ELO are further away
from Reaper during this exact moment, so they do not immediately show a reaction or give a
response. However, there is an expectation that ELO does something so he would not die. This
will also be seen from DEN’s utterance in the ensuing interaction.. TIM later listens to DEN’s
callout at line 7 and starts shooting at Orisa. From this moment onwards, TIM continues
shooting at Orisa until line 10. He then starts casting his ultimate ability then notifies his team
(beating). Overlapping with this, ELO dies in line 10 and reacts to the fact that Reaper was
behind him all along (oh reaper behind). His turn is marked by a change of state token “oh”
(Heritage, 1984), which means that this knowledge that Reaper was behind him has just been
received. Hence, ELO has not shown a relevant second action that would display his
understanding (i.e. running away from Reaper), for he died to the Reaper behind him despite
TIM’s warning in the first pair part back in line 3. After two seconds of silence, which can be
interpreted as an impending trouble (Jefferson, 1989) in line 13 DEN sanctions ELO’s failure
to act in response to the previous warning by TIM as he points out to the earlier turn where TIM
stated that Reaper was behind them (yeah TIM said that). In other words, this turn can be

28
interpreted as DEN’s possible expectation of ELO listening to the warning by TIM and running
away from where he was to avoid dying to the enemy Reaper. Why ELO has not displayed
understanding to TIM’s callout can be explained by looking at TIM’s turn. More than half of
his turn in line 3 is overlapping with DEN’s turn in line 4, thus it might have led to a mishearing
or no hearing by ELO, which can be confirmed by the change of state token “oh” he used in his
turn. Another explanation why this might have happened could be that ELO focusing on Orisa
just like TIM was doing, so he could have missed the callout. Yet, there is no conclusive
evidence for these in the transcript as there was no access to his screen. Furthermore, there was
no claim nor a demonstration of understanding in this transcript. There was however, only a
case of display of non-understanding that led to in-game as well as interactional troubles.
In my data, while usually ignored, such displays of non-understanding are occasionally
followed by sequences in which the player who failed to display understanding is held
accountable for their failure. In other words, the player who does not display their understanding
with a relevant second turn or action gets called out in the talk. These sanctions are carried out
usually at the end of the sequence, and are occasionally remedied with repairs. Next, I will
analyze an excerpt where repair follows the sanctioning. Also, within the same example I will
show a claim of understanding taking place that does not turn into a sufficient display.
In this excerpt, the team is attempting to go to point A in Lunar Horizon in order to deliver an
attack and capture the point. However, DEN and TIM got stranded on the way due to the enemy
team blocking the path while they were attempting to cross earlier. Here, they are re-attempting
to cross the path and unite with their team before attacking. An important aspect of this
phenomenon is “fake go”, which means that the players fake their running action, then return
back to their starting point. They do so to make the enemy team use some of their abilities and
waste them, so they can run back to the spot they have decided without giving any casualties to
the enemy abilities. This can be interpreted similarly to the use of decoys in a real battlefield so
an enemy fires their bullets at the decoy - essentially wasting their bullets.
Excerpt3: Fake go
Map: Horizon Lunar Colony (Assault)
Heroes: TIM (Lucio), ELO (Moira), ATI (Mei), LUC (Reaper), DEN (Orisa)
Enemy heroes: Mei,
1 DEN: er we’ll fake and go here TIM okay?=

29
2 TIM: =yeah.
3 [three (.) two (.) one.
4 DEN: [three (.) two (.) one.
5 ■%✦(3.0) % ✦(0.5)✦
tim ■uses speedboost
tim %runs to gate--%
den ✦runs to gate-->✦runs back and stops
den ✦runs to gate-->
6 DEN: okay they just. ♦ (0.3) ✦not-
mei ♦uses ice wall
den -->✦
7 (0.5)
((DEN’s path is blocked))
8 DEN: #●oh no.-
den ●dies
fig #fig.6
9 (0.2)
figure 6
10 [okay.
11 ELO: [no.
12 LUC: that’s unlucky.
13 (1.0)
14 DEN: i ↑said fake go TIM.
15 and you just $↑left$
16 TIM: ↑hi:h (1.0) $sorry$ (0.5) u:m,
17 (3.0)

30
18 SIN: just go speed him back.
19 (0.5)
20 TIM: yeah (0.8) guess i have to.
DEN opens up by briefing the tactic to TIM that they will “fake and go” in line 1, which
constitutes a first pair part that requires a relevant second action (i.e. TIM to provide an
agreement/disagreement). During the latching turn in line 2, TIM affirms this by saying
(yeah), which can be interpreted as a claim of understanding as the running action has not
started yet. In the overlapping lines 3 and 4, both DEN and TIM do a countdown meaning that
they will go together when the countdown ends. At the end of this countdown, TIM uses his
speed boost ability and they start running towards the gate they want to reach. Until after the 3
seconds silence, TIM continues running and makes it through the gate, he then stops there.
However, DEN has not done the same as TIM, he had started running briefly and went back to
where both TIM and DEN started running from, which was his intention of fake go. After TIM
stops, there is a 0.5 secs silence, after which DEN says (okay they just), which could be
arguably the first indication of a trouble within this action sequence. He tries to cross the path
and reach the gate alive, then enemy Mei uses her ice wall to block his path (marked with ♦,
also see #fig.6). Immediately after this, DEN cuts off most likely due to surprise that mei
blocked his path (△%not-). In what follows, DEN gets killed by the enemy team and says “oh
no” before cutting off again. After a very brief silence (0.2 secs) he says “okay” in line 10,
which overlaps with ELO’s “no”. In line 12, LUC reacts to this by saying “that’s unlucky”.
Similarly to the previous excerpt, in the third position (lines 14-15) DEN addresses TIM’s
failure of providing the relevant second action (i ↑said fake go TIM and you just
$↑left$) with an accusation that TIM just left him alone before he could also cross the gate
himself. He is doing so by referring to his earlier turn in line 1, during which he briefed the plan
to TIM. DEN marks the verb “said” with a rising pitch to emphasize (Walker, 2012) what he
had told TIM earlier, and he repeats parts of that specific turn “fake go” so that he is pointing
out the trouble source. He then finishes off with another rising intonation ($↑left$). In line
16, TIM responds to this first by laughing (↑hi:h), then saying ($sorry$) and produces a
continuation marker “um” (Clark & Tree, 2002). It is also seen that in his turn there are intra-
turn silences in between “hih” and ”sorry” as well as between “sorry” and ”um” (1 sec and 0.5
secs, respectively), which is considered an indication for hesitating (Jefferson, 1989). After
this, a potentially troublesome long silence (3 secs) (Jefferson, 1989) occurs before SIN

31
proposes a candidate solution to the situation (just go speed him back). Meaning that
TIM should run back to the starting location and use his speed boost ability for him so that DEN
can return back to his teammates without spending too much time. Half a second of silence
follows this and TIM affirms this in line 18 by saying “yeah, guess I have to”, and by accepting
the candidate solution, also closes the sequence.
Previously in excerpt 2, we have seen that ELO produced an “oh” prefaced turn that indicated
his late noticing of Reaper and DEN sanctioned this display of non-understanding similarly to
this case. However, there was no response by the sanctioned player (ELO) afterwards, or no
repair attempt took place likely because ELO had died and could not temporarily perform any
in-game actions. This example differs from that in the way that the sanctioned player (TIM)
responded by apologizing, confirming that the understanding did not take place regardless of
the affirmation he had shown in line 2 (yeah). In other words, TIM only claimed understanding
rather than demonstrating it. Moreover, when the sequence was about to end in silence, another
player stepped in to fill in with a candidate repair to handle the situation. TIM carried out the
repair to close the sequence. The excerpts also differed in the way that while in excerpt 2 the
sanctioned player himself (ELO) suffered the in-game consequences of his action (i.e. dying)
whereas in excerpt 3 the sanctioned player (TIM) put another player’s (DEN) avatar at risk, and
eventually it died, which also could arguably explain why more interactional work was needed
to resolve the issue.
From the next sub-chapter and onwards, I am going to present cases where understanding is
actually displayed, rather than merely claimed. I will provide evidence to this from sequential
analyses of players’ in-game actions and their talk.
4.2 Successful Displays
In this section, I will present excerpts where players achieve mutual understanding and
overcome game-related troubles successfully. Three sub-categories will be presented: Verbal
Displays of Understanding, Embodied Displays of Understanding, Embodied and Verbal
Displays of Understanding.

32
4.2.1 Verbal Displays of Understanding
This sub-category focuses on the display of understanding carried out using verbal resources
(i.e. in spoken turns). Display of understanding in verbal turns is a common occurrence in the
data, similar to everyday conversations. How players do this in-game is demonstrated with the
next excerpt. At this moment, the team is attempting to capture the second objective in Temple
of Anubis and end the game by doing so.
Excerpt4: Pull
Map: Temple of Anubis (Assault)
Heroes: TIM (Lucio), ATI (Doomfist), COL (Reaper), DEN (Orisa)
Enemy heroes: Junkrat, Orisa
1 COL: =>do you have pull [do you have pull?<
2 TIM: [amping heal.
3 (1.5)
4 DEN: i’ve in two:
5 (0.5)
6 >i have it now i have it now.<
7 (0.3)
8 COL: >okay okay just tell me just tell me<- [( ) ]
9 TIM: [speeding].=
10 COL: =when you pull them.
As the fight ensues, in line 1 COL quickly asks “do you have pull” in repetition. This is directed
at DEN as COL already knows that DEN’s hero (Orisa) has the pull ability, yet COL does not
know whether the ability is ready or not. Hence, COL’s turn serves as an availability check for
the pull ability. TIM’s turn in line 2 overlaps with COL’s as he is signalling to his team that he
is using healing for his team (amping heal). After a silence in line 3 (1.5 secs), in line 4 it
is seen that DEN responds to COL (i’ve in two:). In this case, DEN is trying to convey
that “I have my ability in two seconds”, this is another case of a shortened expression used by
players commonly. In line 6, he signals that his ability is ready with his quick repetition (>i
have it now i have it now<). COL responds to this in lines 8 and 10 by asking DEN
to let him know when he is going to use his pull ability, in between these lines TIM overlaps
with COL by interjecting his usage of speed boost ability.

33
This excerpt brings evidence to the fact that even in chaotic in-game battle situations, players
are able to display understanding of each other’s turns during their own turns. This took place
between DEN and COL during lines 1-10. COL first asked DEN if he has pull (line 1), which
was attended to by DEN in lines 4 and 6. Then in lines 8 and 10, COL expanded on it by telling
him “okay okay just tell me just tell me when you pull them”. Simultaneously, with this turn,
COL has treated DEN’s verbal display of understanding as accepted. In the end, they created a
successful joint play that led to the capture of the point, with multiple verbal displays of
understanding taking place in the process. Another important note is that COL’s turns explicitly
demand a verbal response. Expressions such as “do you have pull” are information requests
that could be answered with a yes/no, or as DEN did by giving the timing of the ability to
recharge. Furthermore, it creates an increased response relevance in the particular way COL
formed it. First, it is designed as a question that demands an answer. Second, it is repeated twice
quickly to “mobilize” a response from the recipient DEN (Stivers & Rossano, 2010). Hence, it
can be seen that DEN provided the relevant verbal display of understanding. Conversely, if
COL was asking DEN to use an ability that would demand an embodied response (i.e. the usage
of that specific ability), which could be considered as an immediate next step after the check
COL did for the pull ability.
Whereas the previous one was during combat, the next excerpt varies from the previous one in
that it is taken from an out of combat planning situation. Here, the team is organizing their
attack on the first point in King’s Row. They begin by scouting the enemy team from their
starting location, during which they discuss from which angle they can carry out their attack.
Excerpt5: then we go
Map: King’s Row (Hybrid)
Heroes: TIM (Lucio), DEN (Orisa), COL (Genji)
Enemy heroes: Doomfist
1 TIM: they’re holding top right er we should
2 we could rotate behind them or.
3 (0.5)
4 TIM: [through hotel].
5 DEN: [let’s check ]which second dps they have first yeah.
6 (0.5)

34
7 TIM: doom.=
8 DEN: =oh it’s doom yeah yeah we go hotel okay?
9 TIM: yeah.
10 COL: cool cool cool.
TIM tells his team where their enemies are and where his team should attack from in line 1. He
pauses for half a second after he says his turn-final “or”, which may indicate that he is looking
for another candidate location to add to his turn. After the silence, he adds the location in line
4 ([through hotel]), which partially overlaps with DEN’s turn. DEN is making a
contribution to TIM’s plan, which also displays his understanding verbally at the same time
([let’s check ]which second dps they have first yeah). After a half second
silence, TIM displays another verbal understanding by pointing (doom=) (i.e. doomfist) as the
second dps (i.e. damage) hero of the enemy team. DEN, in line 8, marks his turn with the news
receipt “oh” and confirms that TIM’s hotel plan should be carried out (yeah yeah we go
hotel okay?). In lines 9 and 10, TIM first gives an affirmative response “yeah”, followed by
an assessment by COL “cool cool cool”, which closes the sequence. After this, the team then
proceeds to attack from the hotel (lines omitted).
In this excerpt, the players were making an attack plan out of combat so no immediate ability
use or movement was required of them. Hence, it is seen that during this planning moment they
treat the use of verbal resources sufficient and do not make use of embodied resources. The
reason for including this excerpt was to demonstrate the variety of sequences during which
players verbally display understanding. Regularly, the sequences are taken from in-combat
action sequences however the players also talk and discuss out of combat as seen from this
excerpt.
4.2.2 Embodied Displays of Understanding
During this sub-category, the analytical focus is on the embodied resources used to display
understanding of each other’s turns (i.e. gaze, ability use, movement). Close attention will be
paid to the sequential unfolding of the display of understanding moments, similar to earlier
sections.

35
In this excerpt, the team members are standing on top of the balcony in the objective they are
defending, waiting for the enemy team to gather up and attack. Simultaneously, they are trying
to spot where the enemies will approach from.
Excerpt6: Corridor
Map: Oasis (Control)
Heroes: TIM (Lucio), COL (McCree), DEN (Orisa)
1 COL: #they might go corridor↑
tim *>>looks at the main gate-->
fig #fig.7
2 (1.0)
3 or main.
4 (2.0)
5 yea.
6 (1.0)
7 corridor [corridor].
8 DEN: [corridor].
9 *(1.0)*+#
tim -->*.....*looks at the corridor---
tim +shoots-->>
fig #fig.8
figure 7 figure 8
The excerpt starts with COL giving a candidate location in line 1, which is followed by a short
silence (1 sec) and another candidate location in line 3 (or main). These receive no response
during the long silence (2 secs) and the team members continue looking around, walking back
and forth and so on. However, after the line 5 (yea) and a short silence (1 sec), COL spots the

36
enemies in line 7 (corridor [corridor]). DEN also spots the enemies and we see this
due to his turn in line 8 ([corridor]), which overlaps with COL’s repetition of corridor.
Both of these turns can be considered as a first pair part of an adjacency pair, which creates
space for a relevant responsive action. More specifically, because the turn is built as an
announcement of an in-game location, it makes actions related to that in-game location (i.e.
corridor) relevant. These actions can be for instance saying something about the location,
shifting the gaze towards there, shooting there, using abilities targeting that location and so on.
It is seen here that TIM carries out this relevant responsive action. He has been looking at the
main gate since before the excerpt started (#fig.7), which continued until the contributions
by COL and DEN at lines 7 and 8 (see * marker for TIM’s gaze). After TIM receives the
information that enemies are in the corridor, he walks backwards and starts staring and shooting
at the corridor (#fig.8). This shift of gaze and shooting also was done by TIM in a timely
way. He only shifted his gaze and started shooting when his teammates were absolutely sure
that the enemies were approaching from the corridor. The evidence to this can be found by
looking at the multimodal lines between line 1-9. Fundamentally, this gaze shift and the
subsequent shooting action by TIM pictures how players timely use their embodied actions in
Overwatch to display understanding of other players’ turns, as well as their actions. Lastly, the
team was able to successfully defend the point and drive the attacking team away.
In the previous excerpt, TIM displayed his understanding using embodied resources. The
following excerpt will demonstrate variance in the way that all of the team members will
collectively and simultaneously display their understanding using embodied resources.
During this excerpt, the team is preparing to attack point A in Temple of Anubis. They are
preparing to cross the defense enemy team has established at the gates of the point. Their aim
is to make TIM use his speed boost ability, after the use of which they all plan to run together
to bypass the enemy defense.
Excerpt7: Three two one go
Map: Temple of Anubis (Assault)
Heroes: TIM (Lucio), ELO (Moira), COL (Reaper), SIN (Sigma), DEN (Orisa)
Enemy Heroes: Not relevant
Evr - everyone in the team

37
1 SIN: are you ready?
2 (1.0)
3 TIM: i got stunned.
4 (1.0)
5 SIN: [ready?
6 COL: [okay.
7 ELO: yeah yeah.=
8 TIM: =okay three two: one.
9 %■speeding.
evr %runs--->
tim ■uses speed boost
10 (2.0)%#
evr --->%
fig #fig.9
11 DEN: nice.
figure 9
Notice how lines 1-8 serve as a pre-sequence (i.e. lines of talk that prepare a main sequence of
talk/actions to take place) to prepare as a team for the coordinated play (i.e. the joint run), which
we will observe as a synchronized, embodied display from everyone. SIN opens up in line 1 by
asking his team if they are ready. There is a 1 sec silence following this, after which TIM in
line 3 says “i got stunned” to indicate that he is not ready yet. Another 1 sec silence follows this
and in line 5 SIN repeats his question partially (ready?), this overlaps with COL’s “okay”,
which is a minimal contribution to the context. However, it is unclear whether he is just
confirming the situation of TIM, or trying to state that everyone in the team is okay to go. In
line 7, ELO gives the go-ahead response (yeah yeah) to signal that the team is ready. This
latches with TIM’s turn in line 8, during which he counts down (okay three two: one),

38
right before using his speed boost ability and signalling to his team in line 9 “speeding”. The
team starts running slightly before TIM notifies them that he is speeding, nonetheless, they all
are under his speed boost ability’s benefits during the time period. During the silence in line 10
(2 secs), everyone in the team looks at the same direction and runs past the enemy defense as
can be seen (#fig.9), with the exception of ELO, who is not in this particular frame but very
close to TIM around the time where the screenshot was taken. DEN then proceeds to close the
sequence with a third turn (Schegloff, 2007a) with a positive assessment “nice” in line 11.
This excerpt focuses on a recurrent theme in this data where the embodied displays of
understanding demonstrated by every teammate simultaneously. To do this, players frequently
start out with these pre-sequence (lines 1-8) where they plan their upcoming actions, and they
execute this plan afterwards (line 9 onwards), then finalize by confirming the successful display
of understandings (line 10). Players make clear their readiness to make a joint play, they do so
by sequentially placing their talk and actions in relevance to what took place previously. In
other words, before they try to make a timely coordinated play, they ask each other of their in-
game status “are you ready?”, or “are you stunned?” so that they can check if all of the
teammates are ready for it. Their in-game avatars looking in the same direction and moving
towards it in harmony indicates that mutual understanding has taken place and the relevant
actions have been provided by everyone, thus, display of understanding has been done. With
the help of these resources, the players engage in successful joint plays that lead to in-game
success (e.g. capturing the point).
In excerpt 6, we have seen that players make location callouts and look at those directions to
display their embodied understanding, whereas in excerpt 7 it is shown that players are also
carrying out actions such as grouping and moving together while simultaneously looking at the
same direction. Furthermore, in excerpt 6 there was one player who performed the responsive
embodied action while in excerpt 7 there was a pre-sequence which served as a preparation for
the group movement.
4.2.3 Embodied and Verbal Displays of Understanding
Displays of understanding within this sub-category indicate the displays in which verbal
resources are used simultaneously or in contingence with the embodied resources. The
analytical focus will be on the sequential unfolding of these moments as well as the speakers’

39
selection of the upcoming interactant, who will provide a relevant embodied action, and display
their understanding verbally while doing so.
Next excerpt demonstrates how players display understanding towards each other’s turns and
actions using both embodied and verbal resources in contingence. The team is fighting to
capture the first point in King’s Row. They have just confronted their enemies and the battle
ensues.
Excerpt8: Booped
Map: King’s Row (Hybrid)
Heroes: TIM (Lucio), ATI (Mei), COL (Reaper), DEN (Orisa)
Enemy heroes: Lucio, Orisa
1 TIM: △beatin¤g.
tim △uses ultimate
ori ¤uses ultimate
2 they bongo’ed.
3 DEN: okay >♠[i’m pulling] i’m pulling<
4 ATI: ♠[ready]
den ♠uses pull ability
5 DEN: i’m i’m ♣i pulled em TOAste.
ati ♣uses ultimate
6 (1.0)
7 COL: bongo’s one ¤⌂bongo’s one.
luc ¤uses ultimate
col ⌂destroys orisa’s ultimate
8 (0.3)
9 COL: ▼no more bongo.=
col ▼uses wraith
10 TIM: =they [beated].#
11 DEN: [nice ].#
fig #fig.10

40
figure 10
12 DEN: reaper behind >REAPER BEHIND*+
13 #REAPER BEHIND↑< (0.2)+▲ ( )
tim *turns behind
tim +shoots-->
tim -->+▲boops
fig #fig.11
14 (0.5)
figure 11
15 TIM: booped him away.
In lines 1-10 the team can be seen to destroy enemy Orisa’s ultimate ability “bongo” and
announce that the enemy Lucio has used his ultimate ability. DEN then closes this action
sequence by saying “nice” in line 11. He then signals to his team that Reaper is behind them,
which is repeated three times. Two of these are louder and quicker than the first, (reaper
behind >REAPER BEHIND*▲ REAPER BEHIND<). He manages to draw TIM’s attention

41
to the incoming danger. We see this when the second repetition of “Reaper behind” has been
finished, TIM has turned behind to shoot at Reaper (see #fig.10 and #fig.11), which
shows embodied display by TIM. Until the end of third repetition as well as a very brief silence
(0.2 secs), TIM continues shooting at Reaper. After this, he uses his boop ability to push Reaper
away (see line 12), then reports this in line 14 after the 0.5 secs of silence (booped him
away), with which he is displaying his understanding verbally. By doing this, TIM ensures that
his team is safer as there will be no Reaper threat behind his team’s ranks. Consequently, TIM
sets up his team for in-game success (e.g. capturing the first point of King’s Row).
In this excerpt, it is evident that players are competent in conveying information via verbal and
prosodic resources (lines 11-12) and quickly orienting to those using embodied resources (i.e.
TIM’s avatar turning back, shooting, and booping see *, ▲ and + markers in line 12). These
embodied actions show that the information DEN conveyed has been received by TIM, hence
understanding has been displayed with embodied resources initially. However, there is still
space to provide relevant upcoming talk (see the silence in line 13). Line 14 further provides
evidence to this. TIM verbally reports his action after having received the information and
showed his embodied display. To summarize, players are able to make relevant upcoming
actions for each other, conceive these and fill in with the actions required by them in
sequentially relevant positions. While doing so, they effectively display their understanding
using both embodied and verbal resources (lines 11-14).
In the previous example, we have seen that TIM reacted to the calls by DEN (i.e. doing
something to take care of the enemy Reaper) and performed a relevant action (i.e. attacking,
and pushing the enemy Reaper away from the backlines of his team). However, DEN did not
specifically select TIM to do the action there. He only signalled to his whole team, and TIM
picked up the responsibility to fill in with a relevant action himself. This is due to two reasons:
first, as the support player of his team he has to keep his teammates safe. Second, his hero Lucio
has abilities tailored for this purpose (e.g. speed boost/boop). This is similar to a next speaker
self-selection (Sacks, 2004). in a way that TIM self-selected himself to carry on the task.
In the next excerpt, TIM will be selected by DEN to perform a certain action, of which TIM
displays verbal and embodied understanding in contingence. During this excerpt, the enemy
team has won the previous teamfight and TIM’s whole team is waiting to respawn (i.e. avatars’

42
coming back to life) at the next checkpoint. TIM is stranded at the previous checkpoint and is
trying to run back to the next checkpoint where he aims to unite with his team. However, TIM
is not the only person that is stranded. DEN shares his fate, hence he wants to go back to his
team as soon as possible, too.
Excerpt9: Don’t leave me
Map: Blizzard World (Hybrid)
Heroes: TIM (Lucio), DEN (Orisa)
1 TIM: #%okay (0.5) they used [reaper sigma].
tim %runs towards the exit--->
fig #fig.12
figure 12
2 DEN: [don't leave me mitsu].
3 don't leave me mitsu.%
tim --->%stops
tim %runs back towards DEN
4 (0.5)
5 TIM: #got you. ■speeding.
tim ■uses speed boost
fig #fig.13

43
figure 13
In line 1, as regularly seen in my data, TIM starts talking about the enemy team’s ultimate
abilities right after the previous fight has ended. As he is producing his turn, he is already
running towards the exit as can be seen (#fig.12). Before TIM finishes his turn however,
DEN overlaps with him ([dont leave me mitsu]). TIM does not react until after the
second repetition from DEN in line 3. At the end of DEN’s turn in line 3, TIM stops and runs
back towards DEN. Thus, TIM displays through embodied means his understanding of DEN’s
turn. In line 5, he is already united with DEN (#fig.13), and he uses his speed boost while
signalling to DEN that he is going to be safe with TIM (#got you +speeding). By doing
this, he also places a verbal display of understanding because he shows that he has understood
DEN’s turn as a request for help of running back to his teammates. Even though within this
excerpt there is no immediate in-game success, all of the teammates uniting as fast as possible
helps them to engage in successful joint plays. Ensuring teammates grouping up serves as one
way to promote the joint plays.
This excerpt explicates the variance of how interactants fill in with the relevant actions while
displaying understanding both verbally and embodily. Here, DEN selected the next interactant
(TIM) to provide a relevant upcoming action. TIM mainly gets selected due to the hero he plays
(i.e. Lucio), who has the ability to speed boost his teammates so that they run faster. When
compared with the earlier example where TIM turned around and booped Reaper away from
his own team, it is seen here that players use a diverse set of ways to distribute tasks, or pick
them up without explicit assignments.

44
In the next excerpt, the display of understanding is rather reversed to verbal and embodied
rather than embodied and verbal. It is taken from the moment before excerpt 4. The team is
trying to capture the second point in Temple of Anubis. They have just lost AYA, yet they still
want to continue with the attack.
Excerpt10: oh yeah i’m on him
Map: Temple of Anubis (Assault)
Heroes: TIM (Lucio), ATI (Doomfist), COL (Reaper), DEN (Orisa)
Enemy heroes: Junkrat, Orisa
1 DEN: i’m not pulling not yet.=
2 COL: =i think we can still [win this.
3 TIM: [still winnable.
4 (0.5)
5 DEN: sigma l[ow.
6 COL: [orisa low.
7 TIM: orisa no fortify.=
8 DEN: =pulling junk behind us junk behind us.
9 (0.2)
10 ATI: oh yeah i’m on him (0.5) he’s ⌂#one.
ati ⌂kills Junkrat
fig #fig.14
figure 14
11 DEN: nice.=
The excerpt starts with DEN signalling to his team that he won’t use his pull ability just yet
(i’m not pulling not yet). This is immediately followed by COL’s latching turn in

45
line 2. He orients to the fact that AYA died by saying “still”, while showing confidence that
they will succeed in their attack (we can still win this). Overlapping with COL, TIM
does the same in line 3 by telling his team that victory is indeed possible (still winnable).
A negligible short silence (0.5 secs) follows this. Then, in lines 5 and 6, DEN and COL note
that enemy Sigma and Orisa are low, respectively. TIM in line 7 provides more information
regarding the enemy Orisa’s situation by saying she has no fortify ability (orisa no
fortify). DEN’s turn in line 8 starts immediately after TIM’s turn, he is signalling to his
team that he is pulling the enemy hero Junkrat, which is shortened to “junk” to save time. By
producing this first pair part he is not just telling that he is pulling Junkrat here, he is using this
turn to tell his teammates to help him kill Junkrat, hence this turn can be interpreted as a request
for assistance. This is evident from the fact that he is notifying his teammates of Junkrat’s
location (junk behind us) and his action of pulling him. Another, perhaps more important
evidence can be found in ATI’s treatment of DEN’s turn as a request for assistance to kill
Junkrat. After a very short silence (0.2) secs, in line 10, it is seen that ATI is providing the
required assistance to eliminate Junkrat. He receives the new knowledge that junkrat is being
pulled and we see this as he marks his turn with a reactive token “oh yeah” (Young & Lee,
2004). He then signals to his team, and especially to DEN, that he is on Junkrat. A short silence
follows this (0.5 secs) and he conveys the information that Junkrat is close to death (he’s
one). However, even before he finishes his turn, he kills Junkrat. By doing this, he helps his
team capture the point successfully. As with every other killing blow in Overwatch, this
information is conveyed to every player in the game located on the top right corner
(#fig.14). DEN reacts to this in line 11 by saying “nice”, by doing this he treats ATI’s
display of understanding as sufficient and closes the previous sequence where he requested
assistance with Junkrat.
This excerpt demonstrates that during combat situations, displays of understanding can also
occur in a reversed way (i.e. verbal can be first and be followed by an embodied display). It
could be argued that this occurred because in line 8 DEN announced that he required assistance
with Junkrat and that ATI provided verbal display of understanding to this before actually
assisting. Another possible reason for this could be the amount of time the player has had at his
disposal. For instance, when excerpts 8 and 10 are compared, it could be possible that ATI had
more time to react and put a verbal response first to the Junkrat danger in comparison to TIM

46
reacting to Reaper being behind his team, essentially having to use his boop ability first and
foremost, then announcing that he did it.
4.3 Summary of main findings
The data analysis carried out in the current study aimed to explicate the different types of
displays of understanding in Overwatch, and the interactional resources used during these
displays. Two main (i.e. displays of non-understanding, successful displays) categories, three
sub-categories were identified (i.e. verbal display of understanding, embodied display of
understanding, embodied and verbal display of understanding). Making four categories in total,
with each indicating variances in different aspects of social interaction (e.g. repair,
speaker/interactant selection, sequence organization), and they were taken from different types
of in-game interaction (e.g. in-combat, out of combat). In addition, the moments in which
understanding has been displayed successfully lead to more coordinated play within the team
and affected their success positively.
In the display of non-understanding category, within excerpt 1 “Save ults” we have seen that
players can tend to ignore some displays of non-understanding depending on the situation. The
reason for this was that there was no negative in-game consequence for the team. During excerpt
2 “Said that” we have seen that DEN sanctioned ELO’s display of non-understanding as ELO’s
avatar got killed by the enemy Reaper. However, the sanctioning did not lead to further
interactional efforts such as repair to resolve the case, the reason of which could be that the
sanctioned player had already died and could no longer revert the situation. Furthermore, in
excerpt 3 “fake go” DEN sanctioned TIM’s not carrying out the action made relevant by himself
earlier. This, unlike the previous example, consequently triggered a repair attempt by another
team member (SIN). After this, TIM finished the sequence by delivering the relevant
interactional effort (i.e. repair) to solve the problem. From this, it can be seen that during the
first display of non-understanding moment, it was not deemed as a trouble source, hence no
other interactional efforts were made. During the second example, it was indeed treated as a
trouble source by DEN. Nonetheless, repair did not take place. Similarly, in the third example
the display of non-understanding was treated as a trouble source. Yet, repair did take place and
the sequence was thus closed with a problem solution. Moreover, in this category it was also
seen that when players fail to display understanding of each other’s turns and actions, they can

47
be more prone to in-game failures such as the death of their character or failure to capture
objectives.
In the verbal display category, it was seen that players deem it sufficient to put verbal responses
during in and out of combat situations when the verbal response is enough by itself to solve the
situation at hand. Excerpt 4 “pull” brought evidence to this as COL and DEN made a joint play
in combat. Furthermore, the turns designed by COL made it increasingly relevant for DEN to
give a verbal response, during which he simultaneously displayed his understanding of COL’s
turns. Another finding was that TIM’s contributions did not draw any turns addressing them
because they were designed as mere announcements, and were placed in overlaps. In Excerpt 5
“then we go” the players were shown to interact out of combat, during which they deemed it
sufficient to use only verbal resources to display understanding. This was arguably due to that
when out of combat, it is not required of the players to make quick plays with their avatars as
they are not confronted with enemies. Hence, in out of combat situations embodied actions were
rather limited and verbal contributions were more prevalent.
Embodied displays were significant in that the players controlled their in-game avatars timely
coordinated in order to demonstrate that they are in sync with their teammates. TIM’s shift of
gaze and shooting in excerpt 6 “corridor” displayed his embodied understanding of the earlier
turns, where his teammates by their announcements of the said location made actions related to
it relevant. It was also observed that this embodied display was timely placed in order to make
use of in-game time and resources more effectively. While in excerpt 6 it was only one player
that carried out the responsive action, in excerpt 7 “three two one go” it was shown that the
whole team timely and collaboratively carried out actions such as moving. This was evidenced
by their countdown and other pre-sequential availability checks (e.g. “are you ready?”), all of
which served to prepare the team for the timely coordinated play. Another observation was that
players’ avatars looked in the same direction while they carried out this play.
Embodied and verbal displays varied from embodied displays category in the way that verbal
resources were used in contingence with the embodied actions in order to display understanding
of another player’s turn. Within the category, another variance was identified in the selection
of next interactant. During excerpt 8 “booped” we have seen that TIM self-selected himself to
provide a relevant action (e.g. booping Reaper away from his team), whereas during excerpt 9
“don’t leave me” DEN selected the next interactant (TIM) to provide the relevant action. Both

48
of these speaker selections occurred because of TIM’s hero (i.e. Lucio), and its abilities to boop
and speed boost. In the last excerpt of this category “oh yeah i’m on him” it was observed that
the sequential unfolding of displays of understanding was reversed (i.e. to verbal + embodied).
This was arguably due to the amount of time the players had at their disposal. While in excerpt
8, TIM had to immediately react to the Reaper threat behind his team, in excerpt 10 ATI
arguably had more time to first provide a verbal display of understanding to DEN’s assistance
with Junkrat, then deliver the killing blow (i.e. embodied display).
Apart from display of understanding moments, an overarching finding across all of the sub-
categories where understanding was successfully displayed was that players found more in-
game success despite the chaotic, fast-paced in-game battle situations. It could be argued that
this was due to their timely coordinated plays and interactional synchrony. More evidence
supporting this was found as players also carried out actions such as countdowns before timely
coordinated plays, and they engaged in pre-sequential interactions during which they talked
about their upcoming plays before executing them. Players were also shown to make use of in-
game jargon as well as shorthand expressions likely due to the shortage of time and having
quite many tasks at hand to carry out. Another conversation analytic finding was identified in
third position (Schegloff, 2007a). Even though it was not the main scope of this research, it is
worth mentioning because it may contribute to the potential upcoming research in gaming
contexts. Minimal contributions in the third position such as “yeah”, “okay”, and “nice” were
commonly found throughout my data, which are also quite typical in ordinary conversational
settings. However, they could have different implications in game settings.

49
5 Conclusion
The aim of this study was to examine the timely coordinated interaction of players in online
multiplayer gaming settings as the conversation analytic research on this area is still relatively
scarce. Hence, the present study aimed at filling that gap as well. Amongst other reasons such
as popularity of the game, the most important reason Overwatch was chosen as the game to
analyze was that it naturally facilitates timely coordinated interaction with its fast-paced, team-
based qualities. The focus was on displays of understanding within the interaction, for
examining how humans establish mutual understanding is one of the core aims of Conversation
Analysis. Research questions of this study were:
1- What types of display of understanding do participants use in Overwatch?
2- What are the interactional resources used by the participants during display of
understanding moments?
3- How does displaying understanding affect players’ in-game success?
The analysis of the data yielded two main categories (i.e. Displays of Non-Understandings, and
Successful Displays), in total making four types of display of understanding as elaborated on
in the analysis section: Non-Understanding, Verbal, Embodied, Embodied and Verbal. These
categories were based on the most salient features of players’ displaying understanding of each
other’s turns and actions.
Within the first category, examinations were made of moments where players failed to arrive
at a mutual understanding. This was made clear in the interaction through their
misunderstandings, or non-understandings. The players were shown to hold each other
accountable for these moments, and try to overcome these hurdles in interaction by using
interactional resources such as repair. Another observation was that players were more prone
to in-game failures such as the inability to capture an objective when they could not achieve
mutual understanding.
In the second category, moments during which players displayed their understanding through
spoken turns were examined. It was shown that despite the chaotic in-game battle situations,
players were able to arrive at mutual understanding and overcome their enemies. The verbal
turns had prominent features such as quick repetition, or shorthand expressions. It is perhaps

50
due to the fast-paced nature of Overwatch that players interact in this way, which also opens up
the possibility for future research as conversation analytic studies of these interactions remain
largely unexplored in the literature.
In the third category, the examinations were on moments where players displayed
understanding via embodied channels. These embodied channels were mainly based on the
player controlled avatars’ gaze and bodily orientation. It was observed that the players were
able to act in sync and made this available to their teammates through the use of their avatars
for instance by looking at the same place as their teammates or moving together with them.
Moreover, these actions in sync led to more in-game success.
Within the last category, observations were made of instances where both embodied and verbal
channels were used by the players. The players were able to display understanding with their
avatars’ gaze or body movement and support it with spoken turns. This was due to one possible
reason. The in-game roles of the players and the tasks these roles entail make it more relevant
for them to provide announcements after they have made plays (e.g. support character Lucio
booping away Reaper). A reversed sequential unfolding was also observed in this category. In
other words, a player first provided verbal response and supported it with embodied resources.
This was possibly because of the shortage of time and amount of in-game actions creating a
challenge. Another observation was made on the speaker selection, there were instances of both
speaker self-selection and other-selection. This, too, related to the roles of in-game characters
players controlled. Support characters were shown to be more relevant in terms of providing
verbal responses and embodied displays because of their protective and other-enhancing
abilities. As similar to the verbal and embodied categories, the players were able to achieve
better in-game results through their establishment of mutual understanding.
The study context was different from many of the existing research on gaming contexts (see
e.g., Bennerstedt, 2007; 2008; Bennerstedt & Ivarsson, 2010; Mondada, 2013). These research
focused on other types of games (e.g. FIFA, World of Warcraft) in which actions required from
players are arguably slower in pace than a first-person shooter such as Overwatch. Furthermore,
players experience the game take action from a more ‘distanced perspective’, which allows
them to predict upcoming actions with more ease and make their own actions in more
calculated, recognizable ways (Mondada, 2013). Mondada (2013) characterizes this perspective
as a “zenithal view ..., which captures mobile trajectories in a distanced way, maximizing the

51
possibility of projecting and anticipating them” (p. 337). Despite this difference in context,
many of these studies established a common finding that players show their competence of the
game through a set of situated in-game practises (Bennerstedt, 2008; Mondada, 2013; Piirainen-
Marsh & Tainio, 2009; 2014). Players seem to assume that their teammates are competent in
the game, and they constantly engage in competent actions such as displays of attention
(Bennerstedt & Ivarsson, 2010), coordinated teamplay, and visual alignment (Bennerstedt,
2008). The present study contributed to the literature by evidencing in minute-detail players’
competent, collaborative in-game actions, as well as how these are manifested in talk and
action.
As a rather rare study in terms of its context and methodology combined, Rusk & Ståhl (2020)
focused on a first-person shooter game Counter-Strike: Global Offensive (CS:GO) using CA.
Even though they put emphasis on the sequences of Kill (K-) and Death (D-) events, many of
their findings are in line with the current study. First and most central to the present study, they
found that players engage in callouts as part of game strategy (Rusk & Ståhl, 2020). They state
that ”Callouts are employed to co-construct a shared knowledge and understanding of the game
environment through sharing game relevant information such as the locations and intentions of
teammates and opponents.” (Rusk & Ståhl, 2020, pp. 22-23). The present study demonstrated
in detail how players make, and orient to these callouts during talk and action sequences in
gameplay. Second, they illustrated that players in CS:GO engage in efficient communication
(i.e. short and clear interaction) (Rusk & Ståhl, 2020), which resonates with the current findings
about ”shorthand/rapid expressions”. Third, players treat it okay to sometimes not respond to
other players when they are concerned with other immediate in-game actions (Rusk & Ståhl,
2020). Even though this was not highly focused on during the analysis, a similar case was found
in excerpt 2.
This thesis was written in the program of Learning & Educational Technology, and in my study
it was my initial interest to try to understand language learning better. Hence, I find it my
responsibility to elaborate on the language learning aspect of the study. Principally,
Conversation Analysis aims to describe the machinery of human interaction (Sidnell, 2010b).
This is done primarily through spoken and embodied interaction and with a specific interest on
what is made available to all the participants rather than the researcher. Theories are not
consulted and the analyses are based strictly on data and principles of conversation (Seedhouse,
2004). As this is the case, coining an example as “learning” is rather far fetched. Learning itself

52
being understood as an internal, mental, elusive phenomenon, we may need a case to be
repeated a lot of times, or after seeing a learner’s first encounter with a word or phrase (i.e. a
potentially problematic issue for the learner, so we can claim that the learning is possible) we
need to see that word or phrase to be used again in the future to overcome the situation (Kasper,
2004; Wagner, 2004). For this purpose, the research should be designed in a way to acquire
thick longitudinal data. However, even though this can be done, it would still not be certain
whether the future utterances of the same word or phrase is incidental learning or not. In sum,
if learning is understood as an internal mental state, CA might not be able to provide evidence
for other than incidental learning (Wagner, 2004).
On the other hand, if learning is considered from a social viewpoint, CA can provide fruitful
evidence in terms of describing the participation of a learner in the environment (Pekarek
Doehler, 2010). Through the tracking of longitudinal data, CA can show an increase or decrease
in participation, the quality of the participation and so on (Pekarek Doehler, 2010). Building on
this perspective, situated learning can indeed be evidenced and the explanation of inner mental
states can be left for another methodology (Wagner, 2004).
The context for learning can be everywhere, so limiting the analyses to the classroom would be
an outdated fashion (Sundqvist, 2009). Instead, CA’s powerful tools that describe the structure
of language use should be used in everyday contexts in order to understand language learning
better (Wagner, 2004). Following this trend, I was able to describe the language use of a group
of Overwatch players. However, all of the players were already very fluent in their English and
my data was not longitudinal. Because of this, the study does not focus on the initial aim of
shedding light on language learning. To do this, future research should aim to collect much
more data than I could. Future research should also look to recruit participants that are mixed
in their English language proficiency, so that more potential opportunities for language learning
can be created. Another suggestion for data collection is the interface of the participants. In the
present study, the data is from only one participant so accessing all of the participants’ in-game
interfaces was not possible. If the future studies could collect all of the participants’ in-game
interfaces, more detailed and accurate analyses can be made, or different avenues for research
can open up.
In the end, video games form a significant part of everyday life for many people, and with them
human behavior continues to change and evolve. Thus, opportunities for research are definitely

53
rich in the exploration of these new contexts. Not only with CA but also via the use of other
methodologies must we aim to continue developing our understanding of humans in change.

54
References
Aarsand, P. A., & Aronsson, K. (2009). Response cries and other gaming moves—Building
intersubjectivity in gaming. Journal of Pragmatics, 41(8), 1557-1575.
Balaman, U. (2015). A conversation analytic investigation into the impact of task design on
the emergence of information gaps. In Proceedings of 17th International CALL Conference,
Task Design and CALL, Tarragona, Spain (pp. 6-8).
Baldauf-Quilliatre, H., & de Carvajal, I. C. (2015). Is the avatar considered as a participant
by the players? A conversational analysis of multi-player videogames
interactions. PsychNology Journal, 13(2).
Belhiah, H. (2012). Gesture as a resource for intersubjectivity in second-language learning
situations. Classroom Discourse, 4 DOI:10.1080/19463014.2012.671273
Bluemink, J., & Järvelä, S. (2011). Elements of collaborative discussion and shared problem
solving in a voice-enhanced multiplayer game. Journal of Interactive Learning Research,
22(1), 23-50.
Bennerstedt, U. (2007). Avatars & interaction in gaming: Dysfunctional Interaction or a
Practice of Players. In Game in’Action Conference.
Bennerstedt, U. (2008). Sheeping, sapping and avatars-in-action: An in-screen perspective
on online gameplay. In Proceedings of the [player] conference (pp. 28-52).
Bennerstedt, U., & Ivarsson, J. (2010). Knowing the way. managing epistemic topologies in
virtual game worlds. Computer Supported Cooperative Work, 19, 201-230.
doi:10.1007/s10606-010-9109-8
Bennerstedt, U., & Linderoth, J. (2009). The spellbound ones: illuminating everyday
collaborative gaming practices in a MMORPG. In CSCL (1) (pp. 404-413).
Berger, M., Jucker, A., & Locher, M. (2016). Interaction and space in the virtual world of
second life. Journal of Pragmatics, 101, 83-100. doi:10.1016/j.pragma.2016.05.009

55
Caldwell, K. E. H., Osterweil, S., Urbano, C., Tan, P., & Eberhardt, R. (2017). “I Just Don’t
Know Where to Begin”: Designing to Facilitate the Educational Use of Commercial, Off-
the-Shelf Video Games. In Serious games and edutainment applications (pp. 625-648).
Springer, Cham.
Chen, H. H. J., & Huang, W. Y. C. (2010, April). Examining the potentials of computer
games for English learning. In 2010 Third IEEE International Conference on Digital Game
and Intelligent Toy Enhanced Learning (pp. 134-138). IEEE.
Clark, H. H., & Tree, J. E. F. (2002). Using uh and um in spontaneous
speaking. Cognition, 84(1), 73-111.
Drew, P. (2013). Conversation analysis and social action. Journal of Foreign
Languages, 37(3), 2-19.
Ducheneaut, N., & Moore, R. J. (2005). More than just ‘XP’: learning social skills in
massively multiplayer online games. Interactive technology and smart education.
Gee, J. P. (2006). Are video games good for learning?. Nordic Journal of Digital Literacy,
1(03), 172-183.
Goodwin, C. (2000). Action and embodiment within situated human interaction. Journal of
pragmatics, 32(10), 1489-1522.
Hayashi, M., Raymond, G., & Sidnell, J. (2013). Conversational repair and human
understanding: An introduction. In M. Hayashi, G. Raymond & J. Sidnell (Eds.),
Conversational repair and human understanding (pp. 1-40). Cambridge: Cambridge
University Press. doi:10.1017/CBO9780511757464.001 Retrieved from
https://www.cambridge.org/core/books/conversational-repair-and-human-
understanding/conversational-repair-and-human-understanding-an-
introduction/E7ED58DA2A3FA0884CF0F18BD9BDE68D

56
Heritage, J. (1984). A change-of-state token and aspects of its sequential
placement. Structures of social action: Studies in conversation analysis, 299-345.
Heritage, J. (2009). Conversation analysis as social theory. The new Blackwell companion to
social theory, 300-320.
Hindmarsh, J., Reynolds, P., & Dunne, S. (2011). Exhibiting understanding: The body in
apprenticeship. Journal of Pragmatics, 43(2), 489-503.
Hung, A. C. Y. (2007). Video games in context: An ethnographic study of situated meaning-
making practices of Asian immigrant adolescents in New York City. In DiGRA Conference.
Jefferson, G. (1989). A standard Maximum Silence in Conversation. In D. Roger, & P. Bull
(Eds.), Conversation: An Interdisciplinary Perspective. Philadelphia: Multilingual Matters
LTD.
Jefferson, G. (2004). Glossary of transcript symbols with an introduction. Pragmatics and
Beyond New Series, 125, 13-34.
Järvenoja, H., & Järvelä, S. (2009). Emotion control in collaborative learning situations: Do
students regulate emotions evoked by social challenges. British Journal of Educational
Psychology, 79(3), 463-481.
Kasper, G. (2004). Participant orientations in German conversation‐for‐learning. The
Modern Language Journal, 88(4), 551-567.
Keating, E., & Sunakawa, C. (2010). Participation cues: Coordinating activity and
collaboration in complex online gaming worlds. Language in Society, 39(3), 331-356.
doi:10.1017/S0047404510000217
Kohonen-Aho, L., & Vatanen, A. (2020). Participation and co-presence in the virtual world
of Second Life: Transitioning from a gathering to an encounter [Discussion Paper]. Journal
for Media Linguistics. Retrieved from: http://dp.jfml.org/wp-

57
content/uploads/DP002_2020_Kohonen-Ato_Vatanen_Participation-and-co-presence-in-
the-virtual-world-of-Second-Life.pdf
Korkiakangas, T., & Rae, J. (2014). The interactional use of eye-gaze in children with
autism spectrum disorders. Interaction Studies, 15(2), 233-259. doi:10.1075/is.15.2.12kor
Lindwall, O. (2014). The body in medical work and medical training: An introduction.
Discourse Studies, 16(2), 125-129. doi:10.1177/1461445613514671
Lindwall, O., & Ekström, A. (2012). Instruction-in-interaction: The teaching and learning of
a manual skill. Human Studies, 35(1), 27-49.
Manninen, T., & Kujanpää, T. (2005). The hunt for collaborative war gaming-Case:
Battlefield 1942. Game Studies, 5(1), 538-540.
Mondada, L. (2011). Understanding as an embodied, situated and sequential achievement in
interaction. Journal of Pragmatics, 43(2), 542-552.
Mondada, L. (2011b). The situated organisation of directives in French: Imperatives and
action coordination in video games. Nottingham French Studies, 50(2), 19-50.
Mondada, L. (2013). Coordinating mobile action in real time: The timely organisation of
directives in video games. In Interaction and Mobility. Berlin, Boston: De Gruyter.
doi: https://doi.org/10.1515/9783110291278.300
Mondada, L. (2014). Conventions for multimodal transcription. Retrieved from
https://344cc026-c96f-49aa-b4bc-
071b454d3061.filesusr.com/ugd/ba0dbb_986ddd4993a04a57acf20ea06e2b9a34.pdf (last
accessed 22nd April 2020)
Mondada, L. (2014b). Pointing, talk, and the bodies: Reference and joint attention as
embodied interactional achievements. From Gesture in Conversation to Visible Action as
Utterance: Essays in Honor of Adam Kendon, , 95-124.

58
Mondada, L. (2019). Contemporary issues in conversation analysis: Embodiment and
materiality, multimodality and multisensoriality in social interaction. Journal of
Pragmatics, 145, 47-62.
Moore, R., Ducheneaut, N., & Nickell, E. (2007). Doing virtually nothing: Awareness and
accountability in massively multiplayer online worlds. Computer Supported Cooperative
Work, 16, 265-305. doi:10.1007/s10606-006-9021-4
Mori, J., & Hayashi, M. (2006). The achievement of intersubjectivity through embodied
completions: A study of interactions between first and second language speakers. Applied
Linguistics, 27(2), 195-219. doi:10.1093/applin/aml014
Olsson, E. (2012). " Everything I read on the Internet is in English". On the impact of
extramural English on Swedish 16-year-old pupils’ writing proficiency.
Olsson, E. (2016). On the impact of extramural English and CLIL on productive vocabulary.
Paans, C., Onan, E., Molenaar, I., Verhoeven, L., & Segers, E. (2019). How social
challenges affect children’s regulation and assignment quality in hypermedia: a process
mining study. Metacognition and Learning, 14(2), 189-213.
Pekarek Doehler, S. (2010). Conceptual changes and methodological challenges: On
language and learning from a conversation analytic perspective on SLA. In Conceptualising
‘learning’in applied linguistics (pp. 105-126). Palgrave Macmillan, London.
Piirainen-Marsh, A., & Tainio, L. (2009). Other‐repetition as a resource for participation in
the activity of playing a video game. The Modern Language Journal, 93(2), 153-169.
Piirainen-Marsh, A., & Tainio, L. (2014). Asymmetries of knowledge and epistemic change
in social gaming interaction. The Modern Language Journal, 98(4), 1022-1038.
doi:10.1111/modl.12153
Rauniomaa, M., & Keisanen, T. (2012). Two multimodal formats for responding to
requests. Journal of Pragmatics, 44(6-7), 829-842. doi:10.1016/j.pragma.2012.03.003

59
Reeves, S., Brown, B., & Laurier, E. (2009). Experts at play: Understanding skilled
expertise. Games and Culture, 4(3), 205-227. doi:10.1177/1555412009339730
Reeves, S., Greiffenhagen, C., & Laurier, E. (2017). Video gaming as practical
accomplishment: Ethnomethodology, conversation analysis, and play. Topics in Cognitive
Science, 9(2), 308-342.
Rusk, F., & Ståhl, M. (2020). A CA perspective on kills and deaths in Counter-Strike:
Global Offensive video game play. Social Interaction. Video-Based Studies of Human
Sociality, 3(2).
Sacks, H., Schegloff, E.A., & Jefferson, G. (1974). A simplest systematics for the
organization of turn-taking for conversation. Language 50(4), 696-735.
doi:10.1353/lan.1974.0010.
Sacks, H., 1992. Lectures on Conversation [1964–72], 2 Vols.. Basil Blackwell, Oxford.
Sacks, H. (1995). Lectures on conversation. Oxford: Blackwell.
Schegloff, E.A. (1968). Sequencing in Conversational Openings1. American Anthropologist,
70: 1075-1095. doi:10.1525/aa.1968.70.6.02a00030
Sacks, H. (2004). An initial characterization of the organization of speaker turn-taking in
conversation. Pragmatics and beyond new series, 125, 35-42.
Schegloff, E. A. (1987). Some sources of misunderstanding in talk-in-
interaction. Linguistics, 25(1), 201-218. doi:10.1515/ling.1987.25.1.201
Schegloff, E. A. (1992). Repair after next turn: The last structurally provided defense of
intersubjectivity in conversation. American Journal of Sociology, 97(5), 1295-1345.
Retrieved from http://www.jstor.org/stable/2781417
Schegloff, E. A. (2000). When 'others' initiate repair. Applied Linguistics, 21(2), 205-243.
doi:10.1093/applin/21.2.205

60
Schegloff, E. (2007). The adjacency pair as the unit for sequence construction. In Sequence
Organization in Interaction: A Primer in Conversation Analysis (pp. 13-21). Cambridge:
Cambridge University Press. doi:10.1017/CBO9780511791208.003
Schegloff, E. (2007a). Sequence Organization in Interaction: A Primer in Conversation
Analysis. Cambridge: Cambridge University Press. doi:10.1017/CBO9780511791208
Schegloff, E. A., & SACKS, H. (1973). Opening up closings. Semiotica, 8(4), 289-327.
doi:https://doi.org/10.1515/semi.1973.8.4.289
Schegloff, E. A., Jefferson, G., & Sacks, H. (1977). The preference for self-correction in the
organization of repair in conversation. Language, 53(2), 361-382. doi:10.2307/413107
Seedhouse, P. (2004). Conversation analysis methodology. Language Learning, 54, 1-54.
doi:10.1111/j.1467-9922.2004.00268.x
Seedhouse, P. (2005). Conversation analysis and language learning. Language
teaching, 38(4), 165. doi: https://doi.org/10.1017/S0261444805003010
Sidnell, J. (2010). Action and Understanding. In Conversation analysis: An introduction (pp.
59-76). Chichester, U.K.: Wiley-Blackwell. Retrieved from
http://pc124152.oulu.fi:8080/login?url=
Sidnell, J. (2010b). Conversation analysis: An introduction. Chichester, U.K.: Wiley-
Blackwell. Retrieved from http://pc124152.oulu.fi:8080/login?url=
Sidnell, J. (2013). Basic Conversation Analytic Methods. In J. Sidnell & T. Stivers (eds.),
The Handbook of Conversation Analysis (pp. 77-99). Chichester: Wiley-Blackwell.
Sjöblom, B. (2008). Gaming as a situated collaborative practice. Human IT, 9(3), 128-165.
Stivers, T. (2008). Stance, alignment, and affiliation during storytelling: When nodding is a
token of affiliation. Research on Language and Social Interaction, 41(1), 31-57.
doi:10.1080/08351810701691123

61
Stivers, T. (2010). An overview of the question–response system in American English
conversation. Journal of Pragmatics, 42(10), 2772-2781. doi:10.1016/j.pragma.2010.04.011
Stivers, T., Mondada, L., & Steensig, J. (2011). Knowledge, morality and affiliation in
social interaction. In T. Stivers, L. Mondada & J. Steensig (Eds.), The morality of knowledge
in conversation (pp. 3-24). Cambridge: Cambridge University Press.
doi:10.1017/cbo9780511921674.002
Stivers, T., & Rossano, F. (2010). Mobilizing response. Research on Language and social
interaction, 43(1), 3-31.
Sundqvist, P. (2009). Extramural English matters: Out-of-school English and its impact on
Swedish ninth graders' oral proficiency and vocabulary (Doctoral dissertation, Karlstad
University).
Sylvén, L. K., & Sundqvist, P. (2012). Gaming as extramural English L2 learning and L2
proficiency among young learners. ReCALL, 24(3), 302-321.
Tekin, B. S., & Reeves, S. (2017). Ways of spectating: Unravelling spectator participation in
Kinect play. In Proceedings of the 2017 CHI Conference on Human Factors in Computing
Systems (pp. 1558-1570). doi:10.1145/3025453.3025813
ten Have, P. (1997). In the presence of data: conversation-analysis as ‘empirical
philosophy’. In conference on Ethnomethodology, an improbable sociology, Cerisy-la-
Salles, France.
ten Have, P. (2007). Doing conversation analysis; pages 4-13. (2nd ed., ). London: SAGE
Publications, Ltd. doi:10.4135/9781849208895
van de Sande, C. C., & Greeno, J. G. (2012). Achieving alignment of perspectival framings
in problem-solving discourse. Journal of the Learning Sciences, 21(1), 1-44.
Wagner, J. (2004). The Classroom and Beyond. The Modern Language Journal, 88(4), 612-
616. Retrieved June 8, 2020, from www.jstor.org/stable/3588592

62
Walker, G. (2012) Phonetics and prosody in conversation. In: Sidnell, J. and Stivers, T.,
(eds.) The Handbook of Conversation Analysis. Blackwell Handbooks in Linguistics . John
Wiley and Sons, Ltd , Chichester, West Sussex , pp. 455-474. ISBN 9781444332087
Wong, J., & Waring, H. Z. (2010). Conversation analysis and second language pedagogy: A
guide for ESL/EFL teachers. Routledge. doi:10.4324/9780203852347
Wright, T., Boria, E., & Breidenbach, P. (2002). Creative player actions in FPS online video
games: Playing Counter-Strike. Game studies, 2(2), 103-123.
Young, R. F., & Lee, J. (2004). Identifying units in interaction: Reactive tokens in Korean
and English conversations. Journal of Sociolinguistics, 8(3), 380-407.

63
Appendix 1
Transcription conventions
1. Transcription of speech (Jefferson, 2004)
[ A left square bracket on two successive lines indicates the beginning of
overlapping talk by two or more speakers.
] A right square bracket on two successive lines indicates the end of overlapping
talk by two or more speakers.
= Equal signs indicate no pause between the utterances.
(.) A dot in parentheses indicates a pause, which is shorter than 0.2 seconds.
(0.5) Numbers in parentheses indicate the length of a silence in seconds and tenths of
seconds.
. A dot indicates a falling or final indication.
, A comma indicates slightly rising or continuous intonation.
? A question mark indicates rising intonation.
::: Colons indicate the stretch of the immediately prior sound. More colons, the
longer the stretching.
- A hyphen indicates a cut-off of a word.
HI Capital letters indicate a loud voice
↑ The upward arrow indicates rise in pitch.
>hi< Right/left carats indicate that the utterance between them is faster than the
surrounding talk.
ha ha Indicates laughter. Laughter can be referred to in different ways.
( ) Empty parentheses indicate that the transcriber was unable to hear what was said.
(( )) Double parentheses indicate the transcriber’s comment or descriptions.
2. Transcription of embodiment (Mondada, 2014)
* * Gestures and descriptions of embodied actions are delimited between
+ + two identical symbols (one symbol per participant)
Δ Δ and are synchronised with correspondent stretches of talk.
*--> The action described continues across subsequent lines

64
-->* until the same symbol is reached.
>> The action described begins before the excerpt’s beginning.
-->> The action described continues until the excerpt’s end.
…. Action’s preparation.
---- Action’s apex is reached and maintained.
ric Participant doing the embodied action is identified when they are not the speaker.
fig The exact moment at which a screenshot has been taken
# is indicated with a specific sign showing its position within turn at talk.
3. Other abbreviations used in the transcriptions
evr everyone