Content uploaded by Inke Du Bois
Author content
All content in this area was uploaded by Inke Du Bois on Aug 10, 2023
Content may be subject to copyright.
Research Article
Inke Du Bois*
Inside the commentator’s booth: a
Multimodal (Inter)action and Conversation
Analysis on the production of first division
football commentary on TV
https://doi.org/10.1515/mc-2023-0019
Received June 14, 2023; accepted June 15, 2023; published online ▪▪▪
Abstract: This paper discusses the linguistic features of the production of football commentary within its macrosocial
contexts. It contrasts the communication of the TV production team during a football game with the live football
commentary aired on TV. Applying a mixed methods analysis, it reveals how football commentary is prepared and
produced in the technical setting of the commentator’s booth in a stadium. This study reveals how a medial reality of
the same events concerning a goal are broadcasted in the 11 min highlight video football commentary. The author and
her co-principal investigator video recorded a total of 36 h in four work settings. Finally, retrospective interviews with
the commentator and head of sports of the TV station place the study in the broader social context of football as a media
production. In this paper, I show how live football reporting is prepared and conducted using two different theoretical
and methodological frameworks: Conversation Analysis (CA) and Multimodal (Inter)action Analysis (MMIA). Both
provide insight into the synthetic verbal, non-verbal and parallel actions for the successful production of the football
commentary and they also reveal the sequential structure of naturally occurring talk (CA).
Keywords: Workplace communication; TV production; football commentary; multimodal interaction; Conver-
sation Analysis
1 Introduction
“Football on TV is not rocket science”is a statement that was made by a prime time TV commentator in the context
of the present study. Even so creating a shortened video clip of the Bundesliga games aired on prime-time TV on
Saturday nights involves highly skilled and effective work-related team communication. This paper reports on
football as a broadcasting event and media product and illustrates how football commentary on TV is achieved, in
that it employs a practical and theme-oriented approach. The TV broadcasting of football commentary is inex-
tricably embedded in the macrosocial context of its economic value, which will be explicated in the section below.
Currently, live football commentary in the German “ARD Sportschau”has several million viewers, and thus
broadcasting the games requires a high degree of professionalism to attract and retain the TV audience.
Football on German prime time TV is embedded in a network of cooperations and unions with different
production companies. The German Football League (DFL) is a member of the DFB, the German Football Union.
The DFL founded the subsidiary “Sportcast”, a production company, which produces the video footage of all
Premier League Bundesliga games on behalf of the DFL. The DFL is also responsible for the marketing and
broadcasting rights of all football games, for which Sky and the ARD (German Public Television) are contractual
partners. During all the Bundesliga games, Sky provides the live signal of the football games and Sportcast
*Corresponding author: Inke Du Bois, Department of English Speaking Cultures, Faculty 10 Linguistics and Literary Studies, University of
Bremen, Bibliotheksstrasse 1, 28359 Bremen, Germany, E-mail: dubois@uni-bremen.de. https://orcid.org/0000-0002-8113-9826
Multimodal Commun. 2023; ▪▪▪(▪▪▪): 1–20
CORRECTED PROOF
provides the individual images for the ARD for a unilateral production (Steffen Simon, WDR head of sports,
personal interview Cologne, April 1st, 2016). Overall, the Bundesliga is highly profitable; the DFB, the German
Football Union, made a profit of 354 million Euros in 2021, the DFL made a profit of 4.8 billion Euros in 2020 before
the pandemic and lost more than 1.4 billion Euros after the pandemic (Statista 2022). About 48 % of this profitis
made by selling broadcasting rights (DFL Business Report 2022).
My co-investigator Axel Schmid and I joined the TV production team in the stadium of Monchengladbach,
where we set up two cameras each in the commentator’s booth, the “MAZ”(video operator and cutter’s booth in
the stadium, typically in a container), and the directing and editing desks in the OB van (OB stands for outside
broadcasting), as well as in the speaker’s cabin, which happens to be the driver’s cabin of the large OB van. The OB
van is a mobile TV studio in a large truck.
This paper closes a research gap in analyzing how football commentary is co-produced within a complex technical
set-up during the game and how the apparent medial reality of the game in the football commentary differs in
wording, prosody, images, underlining graphics and camera shots. It takes previous research on sports commentary
(Hauser 2009) and documentation of TV production teams (Broth 2009) further in that it combines the research foci
and compares the production process with the finished product. Applying a Multimodal (Inter)action Analysis and a
Conversation Analysis, this approach shows precisely and practically how football commentary is produced from
within the commentator’s booth. It investigates the team communication at the different work stations where the
production of the prime time TV “ARD Sportschau”takes place. This team communication, recorded before, during
and after the football game, ultimately results in the production of the summary of the game video clip.
This paper thus presents a unique comparison between the actual communication during the game versus
the live broadcasted commentary later that night, applying both Conversation Analysis (Sacks 1992; Ten Have
2004) and Multimodal (Inter)action Analysis (Norris 2004, 2019). It uncovers the structured sequentiality of talk,
such as requests and compliances within the team communication and the commentary itself, and highlights
synthetic and parallel actions within the complex technical workplace settings.
2 Theoretical background
2.1 Multiactivity in workplace communication
Workplace communication often includes simultaneous strands of attention such as planning and handling tasks
and communicating with colleagues at the same time (Deppermann 2014). Several studies have focused on the
complex workplace communication within TV production teams and analyzed their communication based on the
sequentiality of talk when multiple media devices are involved in the work environment (Broth 2009; Perry et al.
2019). The studies based their analyses on the sequential organization of talk and the inclusion of visual infor-
mation into CA-based transcription documenting work environments.
The workplace communication in the setting of the TV production team in a football stadium demands that team
members engage in multiple activities. In professional multiparty interaction, “multiactivity is carried out in fast
succession or simultaneously and joint projects are performed within a specific participation framework accom-
plished in situ. Joint projects in this context are organized in terms of changes in participants’bodily formations and
alignments”(Deppermann 2014). Therefore, team communication in the stadium and in the TV studio can include
joint sub-projects such as deciding which camera shots to take for the commentary. Alternatively, the commentator
communicates which cameras will take certain shots to the cutter and video operators, the camera people capture the
images and send them back to the video operators, who then transfer images on the screenbacktothecommentator
for approval. These are projects that areperformedwithintheparticipation framework of the TV football com-
mentary production, which are accomplished ‘in situ’(Goffman 1978, 1981). At the same time, each person in the team
has a very clear participation status (Goffman 1981) relative to the production tasks. Joint projects in this context are
organized in terms of changes in participants’bodily formations and alignments, or in a change of stance towards a
task (Goodwin and Goodwin 2005). While language and communication are limited to sequential organization, the
question is how “multiple streams of activity with different temporal organizations, alternating engagement and
2Du Bois
CORRECTED PROOF
disengagement through embodied performance”canbedisplayedinatranscriptionsystem(Haddingtonetal.2014;
Licoppe and Tuncer 2014). The assumption that participants can engage in several simultaneous actions on differing
levels of attention and awareness led Norris (2004) to the invention of a new transcription system based on the
embodiment of visual non-verbal language, which will be applied in the present study.
2.2 Live football commentary
Live football commentary has been described as its own genre, in that visual actions are narrated at the same time as
they are broadcasted on TV. Seamless streams of actions are connected to culturally wider contexts, and it has been
criticized that there is “no other media sports text that is subject to greater ridicule than the live broadcast (especially
television) commentary which describes for viewers what they are seeing”(Rowe 1999: 98 in Hauser 2009). In live sports
commentary, different rhetorical tools and visually coherent images are applied in mutual dependency to create a
dynamic, emotionalized narrative of the sports event (Hauser 2009: 245). The redundancy which Rowe assumes is rather
a combination of a visual and spoken narrative which dramatically stages the shown images (Du Bois 2022; Hauser
2009). Football commentary is a spoken genre with its own prosodic and syntactic style (Callies and Levin 2019; Gerhardt
2008; Trouvain 2011). For example, sole noun phrases consisting of players’namesormovesmightrepresentaturn.
Thus, verbs or nouns are omitted and very short matter-of- fact statements are typical of football commentary
(Augendre et al. 2018). These unique grammatical omissions and grammatical dislocations are different from everyday
speech. Moreover, play-by-play announcement is characteristic of sports commentary and so is color commentary,
which represents information that is not directly related to the game. Unique para-verbal language, changes in prosody,
pitch and volume, is highly characteristic of sports commentator talk. Research has shown that especially the dramatic
phases of incidents such as goals or fouls are a pivotal aspect of sports announcer talk (Trouvain 2011). Research
demonstrated that not only the commentator, but also whole teams are involved in the production of such a football
commentary. To the knowledge of this author, there is no other study that has analyzed the production process, the
professional jargon of the commentator and his team behind the scenes of the actual game. For TV production teams, it
is crucial that the team members have clear institutional roles with associated tasks and hierarchical positions so that
quick actions can be taken (Raymond and Cashman 2021). A few studies have uncovered the stages of teamwork for the
production of live commentary (Broth 2004, 2009). Mostly, live commentary seems to be spontaneous talk and the
images and commentary aired on TV appear to represent the reality of the football game. However, this paper uncovers
that it is the experienced co-production of many TV professionals that results in the audio-visual media product. This
study contrasts the communication during the live events with the mediated production of the game that is broadcasted.
2.3 Linguistic frameworks
Conversation Analysis (CA), and later discourse analysis, are the first methodological frameworks that provide for
the systematic analysis of the sequentiality of talk. This is of importance, since concepts such as turn (a sequence of
talk), turn transition relevance place (a possible unit of talk where another person might speak), conversational
pairs (e.g. question and answer, request and compliance), interruptions and overlapping talk in conversations
enable the analyst to systematically look at longer stretches of talk. CA traditionally looks at language primarily to
understand contextual information and deducts results from the language data. More recent work within this
framework has also included visual screen shots in the sequential transcription of talk (Goodwin 2000; Heath and
Luff2000; Mondada 2016; Schmidt 2018).
Beyond the linguistic aspects of talk and language being considered a mode, also non-verbal aspects play a
role in communication. A mode is a unit of representation, which “is a semiotic system with rules and regularities
attached to it”. For example, gesture, gaze, posture, proximity and even technical equipment are modes that can
be described systematically (Norris 2011). Gestures, for example, have their own grammar such as (onset, climax
and retraction phase) which synthetically coincide with the mode language, gaze etc. (Stukenbrock 2015). These
simultaneous modes are aspects of communication that can be taken into account when analyzing any
Inside the commentator’s booth 3
CORRECTED PROOF
interaction. In her pioneering work on multimodal interaction based on videorecorded data, Norris builds on the
previous work of scholars who included video data (Goodwin 2000; Kress and Van Leuwen 2001; Scollon and
Scollon 2005) and invents a new transcription system. She includes the above-mentioned non-verbal modes for
analysis, and takes videography as the basis for her transcription system on which she lays language. Multimodal
(Inter)action Analysis is the second methodological framework used for the present study, as it enables an
analysis of modes beyond language such as gaze, posture, proximity as the focus of the transcription (Norris 2004,
2019). Also, modes work together and they represent different levels of materiality, and in this sense the multi-
activity in a workplace setting can be captured more accurately (Norris 2019).
3 Methodology
My co-principal investigator Axel Schmidt and I videotaped the team communication in the stadium and two
weeks later in the studio. On another weekend, we videotaped the studio team in Cologne for a whole 12 h work
day, which ended with the live airing of the ARD Sportschau. In the morning, while we set up our cameras, the
directors, technicians, student bloggers, video operators, audio technicians, and the moderator Reinhold Beck-
mann arrived in the dispatcher room at various points. We videotaped most team communication during the day,
the rehearsal of the different camera positions for the studio and, later on, the rehearsal with the moderator. The
rehearsal included the introductory words for the Bundesliga games and commentaries, as well as back an-
nouncements and transitions for the next video clips of games, and the moderator’s position towards the different
cameras. Finally, we recorded the production of the live show. The Cologne studio is the main hub that co-
produces the short football commentary that is discussed in this paper and is later on imported into the show from
the stadium. There is a variability in the length of the contributions of 8–11 min from the different Bundesliga
games depending on how exciting each game is (Steffen Simon, personal interview 2016). The individual length is
negotiated between the director of the TV show in Cologne that same Saturday with the commentator in the
stadium. We recorded a total of 36 h of video material, which included the live transmission in the studio and at
the director’s, editor’s, audio and graphics technicians’desks, while the live Sports Show was broadcast. Finally,
the author conducted a retrospective interview with the commentator Steffen Simon, who was also the head of
sports broadcasting at the WDR at the time, and the data editor of the ARD, Stefan Knobloch. In the interview with
Steffen Simon, the commentator explicated all the entrepreneurial aspects of the German Bundesliga, the
different communication processes of the team in the stadium, the functions of the screens and he gave insights
into the close team cohesion and hierarchies of the co-workers. I base my analysis on four different sources of the
Monchengladbach stadium data: the first seconds of the game and the event of the goal from the communication
in the commentator’s booth when the game starts, the retrospective interview data, and the first few seconds and
the event of the goal from the live commentary created in the speaker’s booth (Figure 1).
Figure 1: Studio practice in
Cologne: moderator with
camera people.
4Du Bois
CORRECTED PROOF
3.1 Broadcasting entrepreneurship and logistic distribution
3.1.1 Setting in the stadium
In the stadium, the production sites consist of the MAZ (cutters’cabin), the OB van with the director’s and cutters’
desks and screens, which is located in a special area of the stadium parking lot, and the commentary booth, which
is located on top of the audience bleachers (see Figure 2).
All premier league games are equipped with 8–12 cameras including the “hawk eye”cameras that create the
live signal for Sky. ARD’s Sportcast has two further camera people on the sidelines. The 8–12 different camera
angles are transmitted to the screens of the technicians, who select the best angles and shots to create the visual
narrative during the game. All remote TV broadcasting team member communication in the stadium is channeled
through a communication system in the OB van in the parking lot, the node of all communication. Through this
system, the director speaks with the camera people, the Sky editor speaks with the Sky commentator, interviewer
and the moderator, and the ARD commentator speaks with his video operators, field reporters and stadium
director, and with the director on duty in the Master Control room in Cologne.
3.2 Commentator booth-technical settings, layout and functions
In all German premier league football broadcasts, the commentator is the director of production on site (Simon
personal interview 2016). Their ideas about the commentary’s narrative include the selection of scenes and which
visualizations and graphics should be used to emphasize the cohesive sequence of events that are produced in
orchestration with the recording operators or cutters and other team members. They communicate via radio
(intercom). Pressing one button allows them to immediately connect and communicate with the video operator
and cutter, the field reporters and the graphics technicians in the main hub in Cologne.
The left screen in Figure 3 shows the live Sky signal, the right screen shows the current status of the cut (video
material) for the commentary to be aired live in the ARD Sportschau. The commentator speaks his or her text live
onto the images later that night in the speaker’s cabin in the OB van, which happens to be the front of the truck in
the parking lot of the stadium.
Figure 2: Positioning of the
cameras in a stadium (ARD
Documentation Manual).
Inside the commentator’s booth 5
CORRECTED PROOF
3.3 Context of the game
On the day of the game, the researchers joined the commentator and his team for lunch. During the communication at
lunch time the author noted that even though it was hours before the game had started, the commentator and his video
operator were already discussing certain images and narrativestheywereplanningforthefootballcommentarythat
was to be aired later that night on TV. This planning process included the data of Excerpt 1, which is discussed in Section
4 and represents an analysis of the production of the first few seconds of the football commentary. In this case,
background information about the fan groups is provided. The Bundesliga teams 1. FC Cologne and Borussia Mon-
chengladbach are rival teams and due to hooligan raids the year before, only 1,000 Cologne fans were allowed into the
stadium, which led to a protest demonstration by the Cologne fans in the neighborhood Monchengladbach Rheidt. This
story comprised the first images for the football commentary (Figures 4 and5)andwasplannedbeforethegamestarted.
Figures 4 and 5 document the preparation of the visual images for the commentary by the technicians and the
preparation of the spoken text by Steffen Simon.
3.4 Mixed methods: EMCA and MMIA
In order to tackle the analysis of the workplace communication in this TV production setting, I apply two methods. The
sequential nature of language and the synthetic nature of interactions result in a mixed methods approach with the
CA-based transcription system GAT 2 (Selting et al. 2011) and the MMIA transcription system, which combined enable a
holistic view of the production process (Norris 2004). Furthermore, the work environment, the technical equipment,
and also the structure of the communication is dependent on the semiotic technology (Norris 2004; Van Leuwen 2005).
An ethno-methodological/CA-based transcription convention such as GAT 2 focuses on language as the
primary source of analysis. This sequential transcription of language can include para-verbal and non-verbal
actions. Researchers in the field include relevant images within this transcription system to illustrate the
importance of non-verbal aspects in the conversations (Deppermann 2015; Schmidt 2018).
In Multimodal (Inter)action Analysis, the embodiment of any interaction is the basic assumption of multimodal
interaction, and gesture, gaze, head movement and body alignment can be equally used as the prime focus (Norris
Figure 3: Working desk in the
commentator booth.
Figures 4 and 5: OB van planning
of the commentary.
6Du Bois
CORRECTED PROOF
2019). Norris applies an image-based transcription system consisting of panels of screenshots from videos of
conversations. She further extends units of analyses beyond language and introduces the terms “higher and lower-
level actions”(Norris 2019). The terminology of the two frameworks overlaps in that a “turn constructional unit”,a
speaker’s contribution in the terms of Conversation Analysis, such as “hello,”is a “lower-level action”in MMIA. A
pointing gesture or an eyebrow raise also represent a lower-level action, and in this sense, language, gesture or gaze
serve as equal communicative units which are defined as modes (Norris 2019). Since interlocutors’everyday actions
are combinations of these different modes, they are defined as multimodal interaction. The synthetic nature of the
multimodality of the linguistic, para-linguistic (pitch, volume and intonation), and non-verbal modes is included in
the MMIA transcription system which is visually-based, not language-based (Norris 2004, 2020). Thus, unlike other
transcriptions systems, the MMIA transcription system visually captures smaller millisecond increments of actions
based on non-verbal communication. This study uses both methodologies and uncovers the different steps of the
production process, its communication within the team and compares this with the live broadcasted commentary.
4 Findings
4.1 Communication in the commentator’s booth
In the following excerpt, the football commentator who is also the head of sports of the WDR TV station, Steffen Simon
(SI), and the cutter and director’s assistant, Dustin (DU), begin to plan the production of the live commentary of the
Borussia Monchengladbach versus FC Cologne game for theARDSportschau.Theyarelocated in the commentator’s
booth of the Monchengladbach stadium, and, before the game has started, they discuss the sequence of images and clips
that shall precede the game: a simultaneous demonstration taking place in the city of Monchengladbach by the Cologne
fans and the Monchengladbach spectators’protest banner in the stadium. The whole protest is against the reduced quota
ofticketsforColognefans,whichisduetoincidentsofdisorderly conduct that took place one year before during the
same derby. Wolfgang, the stadium director(WR)andlaterMichaelfromMAZ(cutter’s cabin) are present via Intercom
(RA). Steffen speaks into the intercom leaning down to the microphone while looking at the Cologne block. The German
data has been translated into English; the transcription system is GAT 2. Excerpt 1 includes a morpheme-by-morpheme
translation of German for transparency, but for readability this will be left out in the following excerpts.
Excerpt 1:
009 SI: wolfgang,=kannst du uns bitte [einmal den lee]rn:_ähm kölner
name can-2SG you-SG.FAM us-1PL.DAT please once ART.M.ACC.SG empty-PL.ACC cologne-M
Wolfgang, can you please give us the empty Cologne
010 DU: BLOCK auf irgendner kamera geben, die wir uns Abgreifen können,
block on any.one-F.DAT camera give-INF that-F we us-1PL.DAT grip-INF can-3PL
block on some camera that we can pick up on?
011 (1.3)
012 DU: (bist) du eigentlich DRAN,
be-2SG.FAM you-SG.FAM actually it.on
(Are) you actually on it?
013 (0.9)
014 SI: also (.) am liebsten BEIde natürlich.=
PTCL on-DAT nice-SUPERL both naturally
Well, preferably both of course.
015 =der eine ist ja so: SPÄRlich gefüllt,=
ART.M one is-SG PTCL so sparse.like PTCP-fill-PTCP
The first one is so sparsely filled
016 =und der andere GAR nicht.
and the-ART.M other totally not
and the other not at all.
017 (2.7)
018 WR: ja,=aber wir fan(gen)_ja_an den KÖLner senden;=genau.
yes but we start-1PL_PTCL the-M.ACC cologne-M to send-INF exactly
Yeah, but we’re starting to send the Cologne one. Exactly.
019 also (.) ka NEUN wahrscheinlich wirds werden- | oder ZEHN-
so k.nine probably will-FUT.AUX.SG-it becomeINF or ten
So …it’ll probably be K.Or.
Inside the commentator’s booth 7
CORRECTED PROOF
020 (0.7)
021 SI: DANke;
thanks
Thanks.
022 (0.5)
023 SI ka NEUN oder zehn wird dir gleich den BLOCK einmal ablichten.
k.nine or ten will-FUT.AUX.SG you-SG.FAM.DAT shortly
ART.M.ACC block once copy-INF
Kor will capture the block for you in a second.
024 (0.9)
025 RA: ((noise)) DANke; Thanks.
026 (0.5)
027 DU: ok das bedeutet;
Okay this means
The excerpt above represents a request for an action by Steffen and its immediate request realization with an
implied acceptance by the stadium director Wolfgang, who is coordinating the cameras. Steffen requests that the
empty Cologne fan blocks be filmed (lines 9–16). Wolfgang immediately complies with the request by providing
more specific information about which cameras will capture the empty and sparsely filled fan blocks (lines 18 and
19) and he confirms the image sequence that Steffen had in mind. The turn taking is characteristic of institutional
talk in which the participants have an immediate goal (Heritage 2012). The turns are fairly short, smooth and to
the point. Steffen presses the intercom button again as he speaks to the MAZ (video operator and cutter) booth to
inform them about the fact that cameras 9 or 10 will provide the required images for the short clip. In the
commentator’s booth, Steffen’s assistant who is also a MAZ video operator begins the sentence with “ok, this
means”in line 27. Plate 1 represents lines 9–10 of the excerpt in MMIA.
Plate 1: (Translation lines 9 and 10).
8Du Bois
CORRECTED PROOF
The multimodal transcription according to MMIA in Plate 1 (a set of images) is based on video screenshots, which
capture time frames in milliseconds. The multimodal transcription of lines 9 and 10 of the GAT 2 transcript above
begins with Plates 1.1 and 1.2 where Steffen leans his upper body and face to the intercom, with his gaze first in the
direction of the radio and then, in 1.3, in the direction of the Cologne fan block. Thus, his body alignment indicates
that he is focusing on the audio connection to Wolfgang in the OB van at first. In 1.4, his gaze wanders towards the
empty Cologne Block. Both the modes of language and gaze coincide with the words “empty Cologne block”and
then shift in other directions until the gaze fixates on the block in the stadium again after completing the first part
of this request for a camera capturing of the empty fan block. His gaze and words envision the image for the
commentary in 1.8 while his close proximity and turning of the upper body toward the radio remain the same.
Thus, his gaze, head movement and the spoken words “empty Cologne block”indicate his multiactivity, a high
modal density in that the body alignment represents his request communication with the team while his gaze
signals the envisionment of the current empty fan block as an image to be aired on TV later. In the GAT 2 transcript
Excerpt 2, Steffen turns to his assistant Dustin, discussing the sequence of images.
Excerpt 2:
027 DU: okay;=das heiss:t
Okay, that means …
028 SI: =das is_es ZWEIte bil[d?]
That’s the second picture?
029 DU: [ h]m_HM,=
030 SI: JA,
Yeah.
031 (0.5)
032 SI: ähm: (-) und dann:
Um …and then …
033 (0.4)
034 DU: also Erste
So first …
035 ZWEIte bild heisst ERStes bild;=
Second picture means first picture.
036 SI: =JA;=also
Yeah, so …
037 (0.5) 038
DU: DEmo.
Demo.
039 (0.2)
040 (DU:) (mh)
041 (nod nod)
042 SI: erstes (.) (-)
First …
043 erstes bild ist DEmo[:,]
First picture is demo.
044 DU: [ J]A.=
Yeah.
045 SI: =wieviel BILder es sind ist WURSCHT,
How many pictures there are doesn’t matter.
046 und n zweites is DAS,
And a second is that.
047 DU: JA.
Yeah.
048 SI: und dann würd ich eigentlich gerne diREKT ins spiel gehn.
And then I’d actually like to go directly into the game
In this transaction, the assistant Dustin is interrupted by Steffen who completes Dustin’s turn. Steffen states the
sequence of the images he is planning for the commentary, starting in line 28. There is some slight conversational
trouble in lines 34–38, where Dustin initiates a repair about the sequence of images. Steffen Simon continues with
Inside the commentator’s booth 9
CORRECTED PROOF
the planning process not treating Dustin’s comment as a repairable item, but he continues to explain the order of
the three scenes he is planning. In line 46, he reinstanstiates his statement so that Dustin’s other-initiated repair
attempt goes fairly unnoticed as Steffen continues.
Notably, all turns are fairly short and goal-oriented leading to a quick consensus, this is linked to clear
institutional roles and tasks of the interactants. Once a common ground is established in line 43, Steffen repeats
the sequence of the images (line 48) while Dustin agrees with short “yes”in lines 44 and 47 at each statement made
by Steffen.
Using transcription conventions from Multimodal Interaction Analysis of lines 42–48, milli seconds (top left corner
of each single image of the plate) of the conversation are capturedwiththefocusonthechangingmodeofgesture.
Plate 2: (Translation lines 43–48, Excerpt 2).
In Plate 2, the gesture mode serves as a basis for the transcription, together with the view of the stadium from the
commentators’booth. Steffen’s pointing gesture turns into a gesture with his fingers spread into a grabbing or
“taking this image for the future commentary”gesture. Focusing on the mode of body movement, he leans
towards his assistant Dustin and looks at him with raised eyebrows, which emphasizes his statement while he
utters “this is the second image”. In the moment of the request, gaze, upper body alignment, spoken language and
facial expression (eyebrow raise) are activated, in other words, this represents a high modal density. High modal
density means that Steffen is communicating the sequence of images with the illustrated modes in the foreground.
Also, Plate 2 represents a vis-a-vis communicaton inside the commentator’s booth, and the deictic (pointing)
gesture has a double function here: the immediate perceptual space of the empty Cologne fan block and what
Bühler (1934) calls the phantasma (Stukenbrock 2015), a gesture creating the future image (that in this case will be
aired hours later on public TV). The communication continues as follows:
10 Du Bois
CORRECTED PROOF
Excerpt 3:
049 SI: und jetz nich noch DIE mit rein nehmen sondern DIE
And now not bring them in on top of that. Instead actually
eigentlich dann noch holn.=
still get them then.
050 =MÜSsen_ wir wahrscheinlich. (.)
We have to probably.
051 KLAR.
Sure.
052 (1.4)
053 SI: wird zwar bisschen VIEL eigentlich;=n[e,]
Although it’ll be a little much actually, huh?
054 DU: [ J]:A,=ähm: (1.0)
Yes, um …
055 du ich FINde- (.)
you, I think …
056 des KAMman
One can …
057 das kannste auch genau so gut in der HALBzeit machen,
do that in half time as well
058 SI: (.) oder [SO. ]
Or like that.
In line 48, Steffen continues with the statement that he would like to transition directly into the game, using
hedges like “would like”and “actually”, which deviate slightly from the direct requests before. This again leaves a
participant status space for Dustin to suggest the possibility of including the clip of the Gladbach fans in the second
half of the game (line 57).
In a way, he completes commentator Steffen’s turn and slight conversational insecurity (lines 55–58) about
when and how to show an image of the Gladbach fans in the stadium after reporting about the Cologne fans
inside and outside the stadium. Agreeing and interrupting Steffen in line 54, Dustin takes the turn, and suggests
another point in time for the image of the Monchengladbach Fans (Streeck and Jordan 2009).
Plate 3: Excerpt 3, lines 53–57.
Inside the commentator’s booth 11
CORRECTED PROOF
In the multimodal transcript Plate 3, Dustin suggests the sequence of images and topics for the commentary.
After Steffen mentions that reporting on the Monchengladbach fans would be “too much”, Dustin raises his hand
with a “we’ll do that later”waving down gesture towards the Gladbach fans with their protest banners on the
other side of the stadium.
In the next excerpt, the MAZ and video cutters ask Steffen about the length of the Cologne fans’demo in
Gladbach and the empty fan block.
Excerpt 4:
059 DU: [so vo]n WEgen:-=
So sort of …
060 ä[h ] erste halbzeit vorbei-
Uh, first half time over.
061 SI: [JA;]
Yes.
062 DU: stimmung war SCHEISse:,
The mood was shitty.
063 SI: hm[_HM,]
064 DU: [ÄHM;]Um.
065 hier is ne MEInung dazu.
Here is an opinion on that.
066 SI: JA.
Yes.
067 (0.6)
068 SI: JA.
Yeah.
069 ((moves forward with his chair))
070 (1.3)
071 DU: wenn die es denn auf [die KETte kriegen]
If they pull it off…
Notably, during Excerpt 4, the game has not yet started, but the players are only warming up on the field and the
Gladbach mascot foal is cheering up the Gladbach fans. The actual reality of the game (the mascot foal Jünter
cheers up the Mönchengladbach audience) and the mediated reality (mood was shitty) show how the
commentator and his assistant collaboratively design a narrative which is relevant for the TV audience (line 62).
The TV broadcast is a narrative which is created and concentrates on noteworthy events in a sequence that
keeps the tension of the story of the football game at a high level and at the same time not overloading it with
events peripheral to the game.
The sequences of images show firstly how Steffen selects the empty Cologne block as an image that he intends
to use for the commentary, following the outside source video clips of the demonstration of the Cologne fans.
Thus, these analyses reveal that action is often neither peripheral to game play or to the enjoyment of the game as
a spectacle, but it is this constructed experience through the media that provides a look and feel for the edited TV
broadcast that distinguishes it from the “real”(Perry et al. 2019).
Shortly afterward, but still before the game actually starts, the video operator/MAZ (MA) communicates with
Steffen via intercom about having prepared the requested images of the demo and the empty Cologne block:
12 Du Bois
CORRECTED PROOF
Plate 4: Commentator’s booth and cutter’s booth and cutter’s screens below: (Excerpt 5, lines 3–8).
Excerpt 5:
01 MA Also gutes Demomaterial sag ich mal dreißig
Well good material of the demo I would say
02 sekunden?
seconds
03 SI ((gaze on the field)) Dann mach 30 sekunden gutes
05 Demomaterial
Then make it s good demo material
04 MA und fan block wie lange?
And fan block how long?
05 SI Ja dann machst ja noch mal Bild (?) wahrscheinlich 8
06 ja?
Then you are making another image probably yes?
07 MA ((to cutter next to him in the cutter’s booth)) 8
08 Sekunden (…)
s
07 MA Also das Spruchband von Gladbach fällt dann raus
08 oder?
So the banner from Gladbach is gonna fall out then
09 SI Das nehmen wir als HALBzeitbild in Verbindung mit der
We will take that as half time image in connection with
10 FC Geschäftsführung also in sofern merken habt ihr
the FC (Cologne) management so remember that you have
11 über EB bekommen aber am Anfang wär mir das zuviel
received it through EB but in the beginning I would find it too much
12 MA Alles klar (…)
Ok
Inside the commentator’s booth 13
CORRECTED PROOF
The multimodal transcription of line 6 Excerpt 5 above illustrates the different working environments of the video
operator (MAZ) in the cutter’s booth directly below the back view on the journalists (multiple screens and no view on
the actual stadium) in Plate 4 and the commentator booth (screens and the stadium view) whereas they share two
screens as a common ground. Also, it is of note that Steffen states “probably 8”(seconds), indicating the image of the
Gladbach banner should be shown 8 s in the final commentary and whereas the range implied in the probably
coincides with an up and down gesture. Notably, the commentator and the cutter never see each other while they work
together, and the communication takes place via intercom and the two screens (Simon, personal interview 2016). The
right screen shows the state of the art of the cutter’s work. In Plate 4.1, the reader can see the image of the demo of the
frustrated Cologne fans in Monchengladbach outside the stadium, and in the following sequence the images of the
empty Cologne block within the stadium. The final image shows Steffen in lines 9–12 commenting on the position of the
images of the Gladbach fans. Notably, all turns of the MAZ are very short and to the point (lines 1, 4 and 7). This task of
providing images and determining their time frames is smoothly completed in line 12. Finally, Plate 4 also shows the
proxemics of commentator Steffen Simon and his assistant Dustin. Their upper bodies turn closely towards each other
and mostly, while working together in the booth, their body language was synchronous. This reflects their good
working relationship and good communication during the production process. When the researcher made Steffen
aware of the synchronous body languageintheretrospectiveinterview,Steffen states that it really does reflect that
their work relationship is as if he was climbing a high mountain and the other is holding the rope. This metaphor also
relates to the high risk and dependency as their work results, the commentary, will be viewed by up to 6 million people.
4.2 The live commentary
At the end of this day, during the live prime time TV ARD Sportschau on Saturday night, commentator Steffen
Simon is equipped with the bullet points (for the commentary) he prepared during and after the game in Figures 4
and 5 and is now seated in the speaker’s booth with a microphone connecting him with the directors in the OB van
and with the main hub, the ARD studio in Cologne. Excerpt 6 provides a narrative frame around the football game,
a color story before the actual play-by-play announcements begin.
Excerpt 6:
Football Commentary
01 heute vormittag auf den Strassen von MÖNchengladbach rheydt;=
this morning in the streets of Monchengladbach Rheydt
02 =einige hundert Fans des ersten effzeh Köln protesTIEren
some hundred fans of . FC Köln are protesting
03 °h (1) gegen die einschränkung °h des TIcketerwerbes–=
against the limitation of the ticket purchase
04 =norMAlerweise °h gehen Fünftausendvierhundert Derbykarten
05 an die effzeh Anhänger
normally, five thousand four hundred derby tickets are sold to the Cologne fans
06 infolge der ErEIGnisse vor EInem jahr;=
due to the incidents one year ago
07 =als kölner Fans °h das Spielfeld IN Gladbach stürmten,=
when Cologne fans stormed the pitch in Gladbach
08 =wurden heute nur ZweitausendsiebenHUNdert nach Köln vergeben;=
only two thousand seven hundred have been given to Cologne
09 =also nur die HÄlfte.
thus only half of the normal amount
10 °hh diese sind auch noch PersonaliSIERT;=
besides, those (tickets) are personalized
11 =werden also nur gegen die vorlage des PersoNALausweises verkauft–=
so they are sold against presentation of the identity card only
12 =DAgegen wehren sich die kölner: Fans,
that’s what the Cologne fans are fighting against
13 °h besuchen heute (.) nur (gle) gladbacher STRAssen,=
are visiting only the streets of Gladbach today
14 =und meiden weitestgehend das STAdion;
and are mostly avoiding the stadium
15 °h der Gästeblock im Borussiapark ist so leer wie sonst NIE bei einem rheinischen Derby.
the visitor block at Borussiapark is emptier than ever at a Rhenish derby
14 Du Bois
CORRECTED PROOF
16 Gladbach in WEISS,= =köln in ROT–
Gladbach in white Cologne in red
17 °h der effzeh heute mit einer defensiven VIErerkette;=
Cologne today playing with a defensive back four
18 =in der die aussen mit Olkowski und MlaDEnovic °h besetzt sind;=
in which the outside is filled with Olkowski and Mladenovic
Excerpt 6 represents the first seconds of the live spoken commentary Steffen and his team created on that day. In
the live football commentary, the turns are short with unique intonation contours and pauses (lines 2–3) the
grammar exhibits the omission of syntactical units such as Verb phrases (lines 16 and 17). The previous section
showed the complex interactional work and communicative processes put into a few seconds of commentary. In
his live broadcast, as documented in the excerpt above, TV commentator Steffen switches to a sports announcer
talk with play-by-play announcing and matter of fact description of the events in the game (Trouvain 2011). He
thus uses short turns with unique syntactical and intonational patterns such as omissions of verbs, pronouns and
subject verb dislocations (Werner 2019).
4.3 The goal as real and mediated broadcast event
The commentators and his team exhibit expertise and professional knowledge or to put it in other words, their
“historical bodies”(Jones and Norris 2005; Scollon and Scollon 2005) immediately grasp situations and infor-
mation inside and outside the football field and quickly create a coherent visual narrative with the commentator
creating a cohesive spoken text. The historical body is realized through “language in the material world: the
material world is a spatial world, a real material environment full of objects, technologies and signs, upon which
we act semiotically”(Blommaert and Huang 2009). Thus, the commentary is not only filled with the dramatic
phases and excitement, but also with various other narratives, or color commentary. This professional expertise
becomes evident in the event of the goal as it occurs in real time in the commentator booth, and opposed to its
broadcasted product within the live football commentary. The MMIA transcriptions of Plate 5 and 6 illustrate the
contrastive comments on the goal in both environments.
Plate 5: The goal in real time.
Inside the commentator’s booth 15
CORRECTED PROOF
Plate 6: The broadcasted goal.
The comparison of both plates demonstrates that Steffen provides a very detailed analysis of what is
happening before the goal is scored in the broadcasted version while he does not do that in real time. There is a
much higher volume, stronger intonation and a higher pitch when the firstgoalismadebyDahoudinthe
broadcasted commentary compared to the real experienced goal in the commentator’s booth. This is physi-
cally represented in Figures 6 and 7, which are phonetic comparisons of the intonation contours, volume and
pitch, which reflect the differently aroused emotional reactions whilethegoaloccursandwhileitiscom-
mented on.
Figures 6 and 7 show the different formants, the volume, speed and intonation during the event of a goal in
real time and as a mediated production for the TV audience. With a much higher pitch contour, faster speech
and a much higher volume in his voice, Steffen explains the weak defense of the Cologne team. He focusses on
the player Rafael as he passes the ball to Dahoud which results in the one to zero for Monchengladbach. As the
ball reaches close to the goal, Steffen adds the information about the 10 meter hole in the defense which is
Figure 6: Reporting of the goal in
the football commentary.
16 Du Bois
CORRECTED PROOF
visually cohesively supported with a graphic, a white square which illustrates the empty space before the goal.
Additionally, the color comment on the biographical background of the scorer Dahoud is provided simulta-
neously with the close shot on him celebrating the goal in the replay. Steffen’s emotional report is characterized
by a higher pitch, volume and speed which cooccur with the ball getting closer to the goal. His voice thus varies
with the images of the clip that he had preselected and enriched with graphics with his team in the stadium and
in the Studio in Cologne.
5 Conclusions
This paper shows how football on TV is a result of a highly orchestrated art form by many professionals to
produce an interesting, dramatic and cohesive audio-visual narrative. Football certainly is not rocket
science, but the highly-skilled workplace communication of professionals who each have their tasks and
positions within the production team which requires technical know-how. This paper reveals that effective
team communication, professional expertise, clear institutional roles and fast decision processes were the
norm. The TV professionals intrinsically know their position in the team and the communication needs to
run smoothly. Language, gesture, gaze, head movement and proxemics are used effectively in the football
stadium work environment and at the studio in Cologne. The TV production in the football stadium requires
that the interactants are engaged in multiactivity and multiply nested time and space scales, which makes
foreshadowing necessary. Within the goal-oriented production process, especially in the communication
via intercom, the turns are fairly short among team members who are on different production sites due to
the immediate interactional goals. Significant gestures accompany all communication during the produc-
tion, even if other team members do not see them, as they are located in different spaces within the stadium.
However, these gestures indicate the density of multimodal communication, resulting in the intensity of
commands, e.g., in important requests by the commentator to the video operators and cutters for images
that will go into the live broadcast TV commentary. There is a very close dependency of the commentator on
the accountability of his team. The different professional jargon in the football commentary includes pre-
dramatic, dramatic and post-dramatic phases which have different levels of higher pitch, faster speech and
higher volume as was shown in the multimodal transcripts. Further, the commentator switches from a
fairly regular speech style to a typical sports commentator talk with unique grammar (such as omission of
pronouns) and intonation contours. Football commentary is a performance which reflects the commen-
tator’sskilledprofessionalismasheandhisteamproduce the football commentary. Lastly, this study
provides an insight into the embedding of microsociolinguistic and multimodal features within their macro
sociological and economic contexts in which TV production team communication and football commentary
take place.
Figure 7: Commentator’s speech
behind the scenes as goal occurs
in real time.
Inside the commentator’s booth 17
CORRECTED PROOF
Acknowledgments: I would like to thank Steffen Simon, Head of Media of the DFB (German Football
Association), football commentator and head of sports of the WDR at the time of the data collection, for
agreeing to be filmed and interviewed for the purpose of this study. I want to thank Stefan Knobloch, data
editor and script writer at the ARD Sportschau (Working Pool of the Broadcasting Corporations of the
Federal Republic of Germany Sports Show) and the whole team of the ARD Sportschau for their friendly,
down-to earth and humorous manner and willingness to be filmed and interviewed. This project was a
collaboration and joint video data collection with Axel Schmidt, IDS Mannheim, University of Mannheim
and made possible through a funding of the IDS, the Institute for German Language, Mannheim. Co-funding
was provided by Sigrid Norris, University of Wellington (Director of the Multimodal Research Centre AUT
at the time of the data analysis), Meredith Marra, University of Wellington and the Central Research
Funding and Department of Language and Literature of Bremen University. Ulrich Achenbach assisted
with the technical preparation of the video material and Natascha Lüger and Simon Titze assisted with the
transcriptions.
Appendix
Transcription conventions GAT 2
Selting, M., Auer, P., Barth-Weingarten, D., Bergmann, J., Bergmann, P. et al. (2011). A system for transcribing talk-
in-interaction: GAT 2. Translated and adapted for English by E. Couper-Kuhlen & D. Barth-Weingarten.
Gesprächsforschung 12, 1–51, www.gespraechsforschung-online.de/en/2011.html.
[] overlap and simultaneous talk
=latching
(.) micropause (shorter than .s)
(2.85) measured pause geht_s assimilation of words
:, :: segmental lengthening, according to duration
((tie)) non-verbal vocal actions and events
akZENT focal accent
akzEnt secondary stress
?pitch rising to high at end of intonation phrase
,pitch rising to mid at end of intonation phrase
-level pitch at end of intonation phrase
;pitch falling to mid at end of intonation phrase
.pitch falling to low at end of intonation phrase
<<p> > piano, soft
<<f> forte, loud
<<h> > high pitch register
<<p> > piano, soft
<<all> > allegro, fast
°h, °hh inbreath, according to duration
h°, hh° outbreath, according to duration
<<low>> commentaries regarding voice qualities with indication of scope
(such) assumed wording
Transcription conventions Multimodal interaction analysis
Mode: screenshots focus on one mode such as gesture, gaze, head movement, body movement. Higher-level
actions are transcribed with increments of milli-seconds of the specific modes in focus and they are
captured with screenshots of the video. In the example the mode gesture is the focus of analysis
Prosody: represented by wavy font with indicates the tonal frequency
Intensity: represented by letter size and bold font
18 Du Bois
CORRECTED PROOF
References
Augendre, S., Kupść, A., Boyé, G., and Mathon, C. (2018). Live TV sports commentaries: specific syntactic structures and general constraints. In:
The grammar of genres and styles: from discrete to non-discrete units. De Gruyter, Berlin/Boston, pp. 194–218.
Blommaert, J. and Huang, A. (2009). Historical bodies and historical space. J. Appl. Ling. 6: 267.
Broth, M. (2004). The production of a live TV-interview through mediated interaction. In: Sixth international conference on logic and
methodology. SISWO, Groningen.
Broth, M. (2009). Seeing through screens, hearing through speakers: managing distant studio space in television control roominteraction. J.
Pragmat. 41: 1998–2016.
Bühler, K. (1934). Sprachtheorie, 2. Fischer Verlag, Jena.
Callies, M. and Levin, M. (2019). Corpus approaches to the language of sports: texts, media, modalities. Bloomsbury Publishing, London.
Deppermann, A. (2014). Multimodal participation in simultaneous joint projects. In: Haddington, P., Keisanen, T., Mondada, L., and Nevile, M.
(Eds.), Multiactivity in social interaction: Beyond multitasking. Benjamins, Amsterdam, pp. 247–281.
Deppermann, A. (2015). When recipient design fails: egocentric turn-design of instructions in driving school lessons leading to breakdowns of
intersubjectivity. Gesprächsforschung-Online 16: 63–101.
DFL Business Report (2022). <https://media.dfl.de/sites/2/2022/04/DE_DFL_Wirtschaftreport_2022_M.pdf> (Accessed 29 September 2022).
Du Bois, I. (2022). The “goal”: behind the scenes of the „ARD Sportschau“ethnography-of-communication and multimodal interaction
analysis. In: Lecture series. Conference: lecture series sports communication: language, media and culture organized. University of Bremen,
Bremen.
Gerhardt, C. (2008). Turn-by-turn and move-by-move: a multi-modal analysis of live TV football commentary. In: Lavric, E., Pisek, G.,
Skinner, A., and Stadler, W. (Eds.), The linguistics of football. Gunter Narr Verlag, Tübingen, pp. 283–294.
Goffman, E. (1978). Frame analysis: an essay on the organization of experience. Harvard University Press, Boston.
Goffman, E. (1981). Forms of talk. University of Pennsylvania Press, Philadelphia.
Goodwin, C. (2000). Action and embodiment within situated human interaction. J. Pragmat. 32: 1489–1522.
Goodwin, C. and Goodwin, M.H. (2005). Participation. In: A companion to linguistic anthropology. Wiley, New Jersey, pp. 222–244.
Haddington, P., Keisanen, T., Mondada, L., and Nevile, M. (2014). Towards multiactivity as a social and interactional phenomenon. In:
Multiactivity in social interaction: beyond multitasking. Benjamins, Amsterdam, pp. 3–32.
Hauser, S. (2009). Aspekte des Performativen in der Live-Berichterstattung von Sportereignissen. In: Theatralität des sprachlichen Handelns.
Brill Fink, Paderborn, pp. 243–259.
Heath, C. and Luff, P. (2000). Technology in action. Cambridge University Press, Cambridge.
Heritage, J. (2012). Epistemics in action: action formation and territories of knowledge. Res Lang Soc 45: 1–29.1
Jones, R.H. and Norris, S. (2005). Discourse in action: introducing mediated discourse analysis. Routledge, New York.
Kress, G. and Van Leuwen, T. (2001). Multimodal discourse. In: The modes and media of contemporary communication. Edward Arnold, London.
Inside the commentator’s booth 19
CORRECTED PROOF
Licoppe, C. and Tuncer, S. (2014). Attending to summons and putting other activities “on hold”. In: Haddington, P., Keisanen, T., Mondada, L.,
and Nevile, M. (Eds.), Multiactivity in social interaction. Benjamins, Amsterdam, pp. 167–190.
Mondada, L. (2016). Challenges of multimodality: language and the body in social interaction. J. Sociolinguistics 20: 336–366.
Norris, S. (2004). Analyzing multimodal interaction: a methodological framework. Routledge, New York.
Norris, S. (2011). Three hierarchical positions of deictic gesture in relation to spoken language: a multimodal interaction analysis. Vis.
Commun. 10: 129–147.
Norris, S. (2019). Systematically working with multimodal data: research methods in multimodal discourse analysis. John Wiley & Sons, Boston.
Norris, S. (2020). Multimodal theory and methodology: for the analysis of (inter) action and identity. Routledge, New York.
Perry, M., Broth, M., Engström, A., and Juhlin, O. (2019). Visual narrative and temporal relevance: segueing instant replay into live broadcast
TV. Symbolic Interact. 42: 98–126.
Raymond, C.W. and Cashman, H.R. (2021). Institutional roles as interactional achievements: the epistemics of sports commentary. In: Contexts
of co-constructed discourse. Routledge, New York, pp. 23–45.
Rowe, D. (1999). Sport, culture and the media. The unruly trinity. Open University Press, Buckingham/Philadelphia.
Sacks, H. (1992). Lectures on conversation, Vol. I. Malden, Massachusetts, Blackwell.
Schmidt, A. (2018). Prefiguring the future. In: Time in embodied interaction: synchronicity and sequentiality of multimodal resources, 293.
Benjamins, Amsterdam, p. 231.
Scollon, R. and Scollon, S.W. (2005). Lighting the stove. Why habitus isn’t enough. In: A new agenda in (critical) discourse analysis. Benjamins,
Amsterdam, pp. 101–117.
Selting, M., Auer, P., and Barth-Weingarten, D. (2011). A system for transcribing talk-in-interaction: GAT 2. Gesprächsforschung: Online 12:
1–51.
Statista. (2022). Umsatz des Deutschen Fußball Bundes (DFB) in den Jahren 2015 bis 2020, <https://de.statista.com/statistik/daten/studie/
728397/umfrage/dfb-einnahmendeutscher-fussball-bund-gesamt/> (Accessed 29 September 2022).
Streeck, J. and Jordan, S. (2009). Communication as a dynamical self-sustaining system: the importance of time-scales and nested context.
Commun. Theor. 19: 445–464.
Stukenbrock, A. (2015). Deixis in der face-to-face-interaktion. De Gruyter, Berlin.
Ten Have, P. (2004). Understanding qualitative research and ethnomethodology. Sage Publishing, London.
Trouvain, J. (2011). Between excitement and Triumph-live football commentaries in radio versus TV. In: ICPhS. International Phonetic
Association, Hong Kong.
Van Leuwen, T. (2005). An introduction to social semiotics. Routledge, New York.
Werner, V. (2019). A multimodal analysis of football live text commentary. In: Callies,M. and Levin, M. (Eds.), Corpus approaches to the language
of sports: texts, media, modalities. Bloomsbury Academic, London, pp. 183–217.
20 Du Bois
CORRECTED PROOF
