Social Interaction

Abstract

Although mobility has recently gained importance within interactionist studies of social action, not much is known about the consequentiality of being on the move for the particular unfolding of conflictual interactional episodes. Utilising two publicly accessible video clips of ‘road rage’ situations, we describe and analyse the centrality of hand-work in the ebb and flow of emotionally charged interactions between members of traffic. Avoiding an a priori cognitivist stance, we show in detail how the work of hands can be constitutive of anger itself, leading to open conflict on the boundary of physical violence.

1. Introduction

On our daily commutes many of us have directly experienced road rage or witnessed it from a distance. If we have been lucky enough to escape this, we will nonetheless be familiar with it from news media reports of the latest egregious incident of mobile incivility. Many of these reports base their story on videos that have gone ‘viral’ in social media. The coupling of video and news report is not likely to include much by way of social scientific analysis; nevertheless, they contribute to a broader understanding of road rage by helping us to realise that it is no longer just car drivers who may fly into a rage over what their co-movers have done in public life on the road. In short, if you can move about on some vehicle – trucks, buses, bikes, motorbikes, horses, skateboards, e-scooters, monowheels, etc. – then it is likely rage can be found in such mobile interaction. And of course, pedestrians are also able to become enraged by the actions of co-movers.

In this paper, we pick up the utility of social media by using two YouTube videos to analyse conflictual mobile interaction. The first shows conflict between two cyclists, and the second between a cyclist and a motorist. There are growing numbers of detailed naturalistic studies of mobile interaction in various settings (e.g., Haddington et al., 2013; McIlvenny et al., 2014; González-Martínez et al., 2017; De Stefani et al., 2019; Smith, 2021), but the sociological relevance of the topic is most concisely expressed in this recent statement:

Driving as a member of traffic remains fascinating for being ancient yet utterly contemporary, for being locally variable and yet orderly at all points. It is regulated by laws, yet open to improvisation; a site for recognizing and disputing rights; accountable yet with restricted means for seeking and giving reasons to one another for one’s actions. (Laurier et al., 2020, pp. 354-355)

All of this bears reflection; however, it is the last line that most neatly encapsulates the focus for our study. What we will describe are embodied practices and actions that significantly change the course of a piece of mobile interaction. As soon as they are enacted, they are rendered accountable, but exactly as Laurier et al. point out, there are restricted means for seeking and giving reasons to one another for these actions, not least because both parties are (mostly) on the move. Words are exchanged in the pursuit of accountability, with varying degrees of success, but just as Laurier et al. single out the use of horns in interaction on-the-move, we narrow our inquiry to key embodied resources, particularly the use of hands in gesticulating and similar actions. Given that we can replay and slow down video data, and also present it visually via graphic transcripts, it is not hard to reach covering interpretations of the various hand-work moments; nevertheless, we need to be careful in assuming that the members making these actions are working to or from such grounds. This relates to a key point made in the discussion, for as Meyer and Streeck powerfully argue, in studying embodied actions ‘often we are not […] dealing with communication as we used to know it, as a transfer of meanings from one person – or one body – to another’ (2020, p. 324). Rather, we are dealing with an amalgam of rational calculations, bodily reactions, gestalt shifts and unintended effects, all in a material ecology.

Consistent with this, here we work from an ethnomethodological base, focusing more on sequential relations between momentary actions, and how they build into gestalt contextures (Watson, 2008, p. 231; Garfinkel, 2021), providing ‘detailed, concrete observations of organizationally achieved social objects’ (Eisenmann & Lynch, 2021, p. 6), while ‘the tasks of describing the features of cultural events demand a concern for how it’s done, how it’s made, how it comes about’ (Garfinkel, 2019, p. 154, original emphasis). Focus on how takes various forms: it can prioritize identification and description of recurrent specific practices or aim at examination of the overall temporal organization of unique events. Inclined here to the latter, we seek to avoid overly rational, semiotic or mentalistic attributions, and do not assume that hand-work actions are necessarily equivalent to gestures as a form of sign-mediated communication. To be sure, intelligible meaning is detectable in the video data, but the question is how that is contingently and flexibly built up in the dynamism of interactional time, while for the most part being on the move. We begin below by connecting with the existing literature on road rage, particularly Katz’s early (1999) important work. Then we give the source of our data and briefly summarise the events captured in the videos, before moving to the main task of the paper: using graphic transcripts to closely analyse the role of handwork in the conflictual mobile interaction. Finally, we conclude with a comparative overview of our analysis noting the more general relevance of what we have discussed.

2. Some specifics of the ‘road rage’ literature

Both titles of the videos analysed below include a characterisation of the events as “road rage”, illustrating Smith and King’s point that “the phrase ‘road rage’ has come to be part of the social problem landscape for media concern” (2013, p. 478). Not surprisingly then, there is an extensive literature on the topic, some of which expresses discomfort with a straight adoption of the term ‘road rage’ due to imprecision and its sensationalist tone (Smith & King, 2013, p. 479). Frequently, the sociological approaches in the literature spend more time conceptualising and framing rather than attempting to empirically research road rage (Ashbridge et al., 2006; Carroll & Rothe, 2014; Lupton, 1999; Michael, 1998, 2001; Nixon, 2014; Collins, 2008). Or, when empirical research is actually engaged in, it is often reliant on survey or interview data (Balkmar, 2018; Katz, 1999; Lupton, 2002; Nixon, 2014; Smith & King, 2013), which may provide some useful information; however, if the analytical goal is to understand the dynamics of interaction as it is sequentially built up, then such data are limited. Nevertheless, an important point made in much of this work is the identification of the car as a limiting space for interaction. As Nixon summarises, “synonyms found in the literature for the automobile body or shell include ‘carapace’, ‘cage’, ‘cocoon’, ‘exoskeleton’, ‘monad’, and ‘segregated space’” (2014, p. 92). In brief, when something of concern happens outside the car one is travelling in, the very real constraint of nonetheless being in the car, is taken as an important factor in the development of rage.

Katz’s work, even though based on interview data, is one of the best explorations of this theme. ‘Pissed Off in L.A.’ is the first chapter in his book How Emotions Work (1999), and in it he sets off to explore ‘anger as it emerges and declines in social interaction’ (p. 18). Katz alludes to a distinction between the event and its reflection when writing of participants’ characterisations of the scenes that it is ‘often made […] a few moments beyond the heat of the struggle’ (1999, p. 20; emphasis added). Road rage seems to be connected to an attempt to get an apology or another acknowledgement of being right, in the face of the other car ‘driving off as if nothing had ever happened’ (Katz, 1999, p. 22). Katz then establishes the basis for his interpretation of road rage primarily in the aspect of visibility:

Drivers are relatively free to look wherever and however they practically can just because their vision is itself relatively invisible to other drivers […]. For the same reason that the vision of drivers is relatively unencumbered the driver’s ability to speak and, more generally, to express his or her understanding and intentions to other drivers is severely impaired. (Katz, 1999, p. 25)

Confronted with such ‘impairment’, drivers ‘struggle to make the communicative interaction more symmetrical’ (Katz, 1999, p. 26). Apart from using car horns (cf. Laurier et al., 2020) or high beams, Katz notes that ‘hand gestures also acknowledge the inadequacy of sound as a channel of effective expression. It is common to see a driver attempt to deliver a commentary by first maneuvering around other cars in order to get a parallel to an offending driver, and then to launch into some idiosyncratic sign language.’ (1999, p. 27). Deppermann et al. (2018, p. 88) make a similar point, and also note the phenomenon of pursuit of moral accountability over episodes where ‘the overtaken overtakes again and is further overtaken’ (2018, p. 119), though their research has no examples of road rage.

As will become apparent in the description and analysis of our data, gestures do figure in the ongoing development of the two mobile conflicts; consequently, there does seem something to pick up on from this existing literature. We are not disputing Katz’s point about idiosyncratic sign language being resorted to because of the constraints of inter-vehicle communication, however, given that Katz does not have videographic evidence of this, it is actually quite difficult to know whether ‘idiosyncratic’ is the right term. The field of gesture studies (e.g., Kendon, 2004; Kita, 2003; Streeck, 2009) has certainly shown the complexity of gestures in use, nevertheless, researchers are able to venture quite specific glosses of gestures using video or photographic evidence. In other words, Katz’s argument may be expressed in this manner partly due to its methodological limitation: without seeing the gestures we cannot tell whether they are idiosyncratic or widely known and used.

Apart from the existing work of Lloyd on ‘cycle rage’ (2016, 2017), and Smith’s work (2017) which touches upon disputes in cycle lanes, we know of no other research on road rage that has utilised publicly available naturalistic video data. Interestingly, our analysis shows how ‘improved’ conditions of visibility and hearability can also lead to rageful actions. We will explore whether Katz’s observations and interpretations, based on the interactional work of automobile drivers, are also relevant in cyclists’ interactions – including Katz’s fundamental argument that ‘a perception of asymmetrical awareness is a condition of becoming pissed off while driving’ (1999, p. 29), and that the ‘limited expressive possibility in drivers’ social interaction creates an infinite series of ambiguous moments’ (1999, p. 30). This ambiguity is a key argument for Katz, leading to his important claim that road rage is ‘enigmatic’ and ‘tricky to explain’ (1999, p. 20). As he emphasises, sudden elevations in emotions are also quickly dissipated, hence we need to carefully investigate how mobile interactants get ‘taken’ by emotions: ‘What requires explanation is a mysterious metamorphosis. Powerful forces develop against the drivers’ wills, “taking” them against their better judgement. Just as suddenly, the disturbing forces may vanish’ (Katz, 1999, p. 21). Taking into account the temporal sequentiality available in video-recorded interaction, we examine how such ‘ambiguous moments’ where control is lost and people are taken by strong emotions, are produced and rendered as meaningful by the road users in their ongoing interaction in real time. Our focus on hand actions and gestural work is not because we are gesture scholars, nor want to make a key contribution to that field¹, rather this focus stems from the empirical material itself. In watching it, the hand and gestural work seems inescapably important, and our key concern is to analyse its part in the sudden rise, and fall, of heightened emotions between interactants on-the-move.

3. Empirical data source, summary of the events

The materials analysed in this article consist of two video recordings ‘produced by observers of isolated incidents for purposes other than to serve as research data’ (Whitehead et al., 2018, p. 331) and made available on the video-streaming website YouTube. The first video is 2 minutes 13 seconds long and is titled Cyclist vs Cyclist Road Rage. It was first posted by ‘CBL’ in May 2017, and at the time of writing is still available at https://www.youtube.com/watch?v=9N5wibjmDA8. For the sake of brevity, below we will refer to this via the acronym CvC, and we will refer to the two cyclists as Camera and Beard: Camera is the filming cyclist, and Beard is the cyclist with the significant beard. Although the YouTube metadata does not provide geographic information, we were able to identify the site of the event as London, and even locate the specific streets by using Google Earth. It is thus possible to have a virtual ‘walk through’ the relevant areas and get a better sense of the place and its proportions. The recorded incident starts next to the McDonald’s on Evelyn Street, the verbal confrontation happens after the turn left to Bestwood Street, and the video clip finishes after a significant amount of riding (which is cut out from the YouTube video) at the crossing of Jamaica Road and St James’s Road.

The second YouTube video is 1 minute 51 seconds long and is titled Kingston cyclist defends himself with D-lock in road rage attack. It was posted by ‘Surrey Suburban Cyclist’ in August 2016, and at the time of writing is still available at https://www.youtube.com/watch?v=nXnTvScGnCc. We will refer to this via the acronym CvD, and we will name the main protagonists Cyclist and Driver, for obvious reasons. Similarly to the first clip, we were able to specify the location of filming as Elm Road, ending at the corner with Florence Road, in Kingston upon Thames, a suburb of London.

Given that both of the videos used for data are publicly available, and neither resulted in any police or legal intervention, in the screensnaps provided in the analytic section below we do not anonymise the participants via blurring, pixelation or grey scale techniques. This conforms to the International Visual Sociology Association’s (IVSA) code of research ethics (see Papademis & IVSA 2009). Additionally, it is important for our analysis that we can clearly see the details of faces, gaze direction and so on. Below we offer quick textual glosses of the two videos in order to provide the reader with a preliminary overall comprehension of the events before we move to detailed analysis ‘bit by bit’ (Goodwin, 2018, p. 48). This initial summary is no substitute for viewing the videos but serves to provide the gist of what happened in the two episodes of mobile interaction. Also note that in the analysis below the form of our graphic transcripts is relatively simple; we do not try to incorporate the kind of detail often found in conversation-analytic approaches to multimodal interaction (see Laurier, 2019 for a recent discussion of the adequacy of relatively simple transcription).²

3.1 CvC summary description

Camera is cycling on a busy street and when a large truck stops ahead of him, he waits behind the truck instead of trying to ride the narrow gap between the truck and footpath. As he is stopped, he hears ‘come on mate’ from Beard behind him. As the truck moves off both cyclists ride in an unoccupied bus lane, and as Beard passes Camera he says, ‘if you’re scared you shouldn’t be on the fucking road’. Camera quickly responds, perhaps somewhat startled, by explaining that his stopping was perfectly sensible as moving inside trucks is ‘how people have died’. However, Beard, who has moved ahead of Camera, simply responds by taking his right hand off his handlebars and making a ‘wanker’ hand action. This insult enervates Camera to respond, ‘I’m a wanker because I don’t want to die down the side of a truck?’ Beard does not respond but moves further ahead. Shortly thereafter, both have to stop behind a bus and other traffic, giving Camera another chance to engage with Beard, who nevertheless continues to say nothing. At this point there is an edited break in the video (which seems to be as much as a 10-minute ride, according to Google Maps) and we next see Camera moving to stop at a red traffic light where Beard has already stopped. This gives him another chance to take up his case, and whereas Beard simply responds with a stony stare and two exclamations of ‘fuck off’, Camera has a lot to say, including an invitation for Beard to ‘say to the camera’ his previous claim about a ‘scared wanker’. This verbal exchange while stopped at the traffic light ends with Camera moving off in front of Beard, exclaiming ‘shut ya fucking mouth!’.

After about another 75 metres of cycling, we see Beard ride up alongside Camera, and now with one or both hands off his handlebars, attempt to enlarge on why he felt Camera did not need to stop behind the truck. Camera is unconvinced. The video ends with Beard riding in front of Camera, with the latter emphasising the difference in risk between riding inside a car and a truck, finally exclaiming, ‘you’ll find out one day when you’re under it’. There is no physical conflict between the two cyclists, but over the course of the reasonable distance they travel we witness a significant dispute, including derogatory name-calling and resort to swearing (both typical in bluster confrontations – see Collins, 2008). When the mobile interaction ends, there is clearly still a significant difference of opinion, and of course Camera is sufficiently moved to later post the video of the incident on YouTube.

3.2 CvD summary description

Cyclist is moving along a suburban single carriageway (a 20km zone), with car parking on both sides of the road. About 15 metres ahead and to the right, a sports car with a sole occupant pulls out into the road. There is nothing that appears particularly dangerous about this manoeuvre, nevertheless, when he is alongside, Cyclist reaches out and taps the left-wing mirror of the car. Immediately, Driver looks back at Cyclist, raising and pointing his left index finger at Cyclist, who quickly responds, ‘no, no, fuck off’. Here, Cyclist gets ahead of Driver, but the latter’s car can be heard revving and we then see him pass and move to the left, in what looks like an attempt to stop Cyclist. Cyclist does not stop, but yells out, ‘you’re on camera so fuck off’; in the background to this we hear Driver’s ‘you’re a fucking idiot’. Then around a bend in the road, where it reverts to a two-way carriageway, Driver catches up and comes close to Cyclist, succeeding in stopping him (we hear a squeal of the bike’s brakes). This time, it is Cyclist who makes an index-finger point, simultaneously exclaiming, ‘you pulled over in front of me, it’s all on camera’ (we can also hear significant heavy breathing from the Cyclist at this moment).

Immediately, Driver gets out of the car and as he comes towards Cyclist around the back of the car, he has his right hand clenched to make a fist. His whole body posture is agitated, and he repeatedly exclaims ‘you’re a fucking idiot’. Despite this agitation – what looks to be a clear desire to hit Cyclist – he backs off. Perhaps ‘confrontational tension’ (Collins, 2008) is acting as a brake on his inclination here, but there is another strong limiting factor in operation, and this is apparent in other frames of the video: Cyclist has taken his metal D-lock off his bike and is holding it in front of his body in a defensive manner (he also exclaims ‘get back! get back!’). Thwarted, Driver returns to seat himself in the car, from where there is a longer attempt to give his version: in short, he did nothing wrong but ‘kept right over to the right-hand side’. Cyclist still holds to his version that Driver was too close to him, and so the dialogue takes the form of repeated gainsaying. Finally, Driver gets out of his car a second time, this time to check where Cyclist had tapped the wing-mirror. Seeing no damage, he gets back in the car, simultaneously returning to repeated exclamations of ‘you’re a fucking idiot’. Cyclist’s rejoinder to that is ‘watch the video on YouTube later mate’.

2. Analysis

2.1. CvC: ‘You’re a scared wanker’

We are going to use four graphic transcripts to show the key embodied work, including gestures, in this conflictual interaction. We also draw attention to the importance of the spatial position from which it occurs, specifically whether it is from behind, alongside or in front of the other cyclist. Figure 1, panels 1 to 3, show that the first interaction between Camera and Beard occurs without visual contact. Stopped waiting for the truck to move, Camera hears from behind a short voiced complaint – ‘come on mate’ – from Beard, who may have stopped but at the very least cannot proceed around Camera. Similar to other encounters in public space, the opening has the form of ‘articulating the reason for the approach (rather than greetings)’ (De Stefani & Mondada, 2018, p. 268), in this case the possibly inadequate behaviour of stopping rather than riding, thus creating a miniature traffic jam. As Camera moves off behind the truck and moves into the bus lane, he makes a justificatory defence of his cautionary stop: ‘I’m not going down the side of that, mate – you want me to die’. It is made clear that his stopping was for cautionary reasons, with the consequences being expressed in an exaggerated form – you could die riding down the side of the truck. Panel 4 shows, via the silhouette, that as Beard rides alongside, Camera is looking at him. Doing so, he receives an intensification of the complaint: the negative judgment ‘if you’re scared’ is made, and the introduction of swearing in ‘fucking road’ intensifies Beard’s negative judgement of Camera. Note that in this initial exchange both cyclists communicate with their hands in what we can call ‘home position’ (cf. Sacks & Schegloff, 2002), that is, with both hands on the handlebars. This position, in which the whole cyclist’s body is in operative unity with the bike, observably indicates that the principal activity of Camera and Beard is riding, although they simultaneously partake in talk.

Focusing on panels 4 and 5, we can note a contrast between this mobile interaction and how complaints are typically handled in stationary interaction, for instance in F-formations (Kendon, 1976). Having made the complaint and the negative judgment, Beard does not wait alongside for a response but continues to move ahead. It is then from several metres behind that Camera essentially repeats his earlier justification, this time adding his own swearing amplification: ‘that’s how fucking people have died, yeh’. Adapting Deppermann et al. (2018), it seems fair to say here that Camera regards Beard’s complaint as ‘illegitimate’, perhaps framing Beard’s actions as self-directed, making him akin to what they call ‘egoistic drivers’ (2018, p. 118). In this context, panel 6 shows what we think is a very significant action. Steadfastly looking and moving ahead, turning his head slightly right, Beard takes his right hand off his handlebars, holds his hand in a loose fist and moves it up and down. There is no doubt about the intended interpretation: Camera immediately responds with, ‘I’m a wanker because I don’t wanna die (.) down the side of a truck’ to Beard’s back. In gesture studies, this hand motion is known as an ‘emblematic gesture’, that is, it is part of a culturally shared inventory that is ‘quotable’ (Kendon, 1997, p. 118). Summarising this instant very well, Kendon states that ‘they serve mainly to convey messages of interpersonal control (orders, commands, threats) as a component of an interaction ritual, or as evaluative expressions of the personal state of the self or of others’ (1997, p. 118). In short, Beard insults Camera, without words, via the easily understood ‘wanker’ gesture.

Katz’s identification of ‘asymmetrical awareness’ as a key factor in the genesis of road rage, thus seems very applicable here. Beard acts as if making a complaint about something that Camera holds to be self-evidently justified, and calling him scared, and gesturing that he is a wanker, were all reasonable things to do. Clearly, for Camera they are not – both Beard’s verbal expressions and gesturing are illegitimate – but the spatial problem he has is that by now Beard is too far ahead to be in a participation framework. This departure from what we could call the scene of insult no doubt helps to ramp up Camera’s emotional state. It is also the beginning of a rather long ‘continuing state of incipient talk’ (Schegloff & Sacks, 1973), which is tied to the retrospective accounts and formulations of this initial incident.

Not shown in a graphic transcript is the next 30 seconds of the video, but it is easily summarised. Fifteen seconds after the last panel of Figure 1, Beard is forced to slow down and then stop behind a bus in the bus lane (vehicles are also stopped in the main carriageway). Camera takes this opportunity to pull up alongside him, where he says, ‘I don’t think you quite realise how dangerous that is, mate ((Beard begins riding around the bus)) yeh, you just don’t give a fuck, do you ((said to Beard’s back))’. Then, as noted in the summary description, at 58 seconds there is an edited break in the video. Figure 2 shows the first significant development from the recommencement of the video after the edited break.

At one minute into the video, we see Camera pull up at a red traffic light. Beard is already stopped there, looking ahead, and before Beard sees Camera alongside he takes his hands off his handlebars placing them on his thighs, as we see in panel 1 of Figure 2. As Camera says, ‘Can you just do me one favour, mate’ he maintains the body position but now looks toward Camera. However, a slow replay of the video shows that as Camera is alongside, and throughout his utterance, Beard also moves his left leg slightly up and down in a pumping motion. It is the kind of body action associated with agitation. This seems a fair interpretation because as shown in panels 2 and 3, as Camera elaborates that the favour he is asking is ‘just one thing’, he is also holding up his left index finger, in a ‘listen to me, I have one important thing to say’ type gesture. Effectively, Camera is ‘lecturing’ Beard about his unreasonable actions: he rejoins his framing of Beard’s actions as ‘illegitimate’, asking for his complaint to be repeated to ‘the camera’, which is simultaneously pointed to by his left index finger. Moreover, the ‘scared wanker’ categorisation that had previously been directed to Camera, is now broadened out to a general category ‘you’re a scared wanker’, that is, any cyclist not taking that route down the side of a ‘tipper truck’ qualifies for the insult of being a ‘scared wanker’. In some ways then, by asking for Beard’s complaint to be repeated to the camera and by invoking general categorisations of action – that is something a ‘scared wanker’ would do – Camera is enrolling larger populations in his counter-complaint to Beard.³ Or at least, his attempt to get his counter-complaint heard.

Here, Camera seems to be looking for an apology, or at least a recognition that Beard’s complaint went too far and was illegitimate. As shown in Figure 2, Beard’s response, fascinating in its detail, both verbal and embodied, is short and unequivocal. Maintaining his upright body position, he simply says ‘fuck off’. Note here that because he is wearing sunglasses, any possible emotional reaction visible in the eyes is unavailable to Camera. Then, as panels 6 to 7 show, Beard leans forward and moves his hands to home position on handlebars, orienting to riding as the principal activity while perhaps also indicating some awareness of the more emotionally charged state the interaction has evolved into, thus wanting to be ready to go when the light turns green. The second ‘fuck off’ captured in panel 7 is also uttered with this time a slight but perceptible rocking up and down of his head. From here, this time it is Camera who moves off first, talking back to Beard in a series of intermittent ‘body torques’ (Schegloff, 1998), significantly ramping up his counter-complaint, as shown in Figure 3. The gist is that Beard is welcome to ride in a dangerous manner, but given that Camera doesn’t want to be exposed to such perils, Beard should ‘shut ya fucking mouth’. Just captured in the video, as the last forceful statement is made, is Camera’s left hand engaged in some kind of gesture, perhaps another ‘emblematic gesture’ like a closed and shaken fist.

Camera, as the overtaken who was abused, has now overtaken the abusive overtaker, and is out in front, having made clear at the stop his counter-response to Beard’s complaint of having held him up. He rides out in front for 10 seconds, then Beard comes into view alongside him, thus constituting a ‘togethering’ (Ryave & Schenkein, 1974, p. 269), where they enter into the longest passage of dialogue (spanning 45 seconds) they have. It is not always clearly audible, but one of the other interesting aspects of it is the centrality of gesticulation, as shown in Figure 4.

Panel one shows that to begin with, Beard keeps one hand on his handlebar as he says, ‘The surprising way to keep up down there’, but almost as soon as he starts talking to Camera he takes both hands off his handlebars, also clearly looking leftwards to Camera in what we have to assume is mutual gaze. Quite apart from what he communicates via talk and gestures, Beard here is also showing his masterly competence as a cyclist. Most competent cyclists can ride in this both-hands-off manner; nevertheless, to do so now clearly instantiates and embodies this competence; moreover, it also indicates a willingness to take a risk. That is, riding with both hands off, while not particularly difficult, does come with the risk that if a bump in the road is suddenly encountered, the cyclist can crash onto the road. Looking back on the actions of Beard, we can now see this willingness to take risks on the road has been a central theme of his interaction, even though minimal, and in effect now it is crystallised in both what he says and how he employs his whole body and hands in these few crucial seconds of interaction.

Despite this clear attempt to engage with Camera in a much more amicable manner – note there is no swearing – Camera rejects Beard’s position. After Panel 3, while Beard is looking at him, Camera immediately responds with, ‘the thing is (.) I’m not in a hurry (.) I’m not in a hurry to die’, repeating in a different way his suggestion to Beard at the stop to ‘have a little patience in your life’. Upon hearing this, it is with another hand gesture that Beard shows he is uncertain about Camera’s suggestion. Panel 4 shows him holding both hands off the handlebars, in a palm-up, open-hand gesture (Müller, 2004) which indicates ‘absence of knowledge’ (Cooperrider et al., 2018) and acts to ‘strongly select the recipient as the next speaker’ (Reynolds, n.d.). Thereafter there is more dialogue around their differences; nevertheless, neither rider reaches agreement with the other, and in effect Camera ends with a warning that Beard’s riding style will find him under a truck one day. The last we see after this is Beard riding off in front.

2.2. CvD: ‘You’re a fucking idiot’

In the second case, there is a more rapid move to confrontational actions with a different kind of gestural work involved. As the initial description above established, this second case significantly differs from the first, as it involves a cyclist and a car driver, and it also comes much closer to physical violence (beyond ‘mock aggression’, see Afshari Saleh, 2020). Similar to CvC, the CvD video is captured from an action camera on Cyclist’s helmet, though it is mounted off-centre to the right. We also need to stress here that, as is often the case with videos recorded whilst cycling, there is significant ‘windage’ noise, meaning that sometimes utterances and other sounds are unclear or inaudible. Figure 5 shows screensnaps from 7 to 27 seconds of the video.

Panel 1 shows Cyclist riding with hands in ‘home position’; that is, both on handlebars. From panel 2 to 3 we see that as the silver sports car pulls out into the main carriageway, Cyclist reaches out his right hand and taps the left wing mirror of the car (as far as we can determine, just the once). Interestingly, we can also see that the wing mirror is taped up. From common knowledge, we know that car wing mirrors frequently get knocked or broken, and if the damage is not too severe they can be repaired by taping them up. Of course, we cannot definitively say that the Cyclist noticed this; however, the question could be asked if the subsequent course of the interaction would have been different if noticing this had led him not to tap the wing mirror. For, apart from the taping of the wing mirror, the car is noticeably very clean, tidy and undamaged, so the selection of that particular place to tap may be significant for Driver. Undoubtedly consequential is that the car has its top down, thus enabling Driver to hear Cyclist tapping the wing mirror. As panel 4 shows, after the tap he turns his body, takes one hand off the steering wheel and uses the freed hand (panel 5) to point directly at Cyclist. Simultaneously with this point, Driver does say something, which unfortunately is inaudible. Nevertheless, this strip of interaction is not hard to interpret, because of both the hand gesture itself, and Cyclist’s next response being a clearly audible ‘no, no, fuck off’ (panel 6). The latter action also serves to emphasise the reciprocity of visibility afforded by the open-top sports car, that is, Cyclist can easily see Driver’s finger point, turned body and facial expression, and also hear what he utters. Simultaneously, Driver makes a noticeable veer left towards the cyclist which seems designed to stop him so he can press his complaint. It is important to note here that from the point of first physical contact – Cyclist’s tap on the wing mirror (panel 3) – there is only two seconds until Driver moves his car left in an apparent attempt to stop Cyclist (panels 4 and 5). This is a remarkably short period of time, within which very significant social actions are occurring.

Here we can usefully draw upon Goodwin’s work (2003) and his emphasis on the importance of ‘activity frameworks’. An activity framework specifies which features of the environment are relevant for the ongoing activity and hence are likely to be the referent of a pointing gesture. In other words, as Watson (2005) has aptly put it, public space is an ‘environment of normal appearances’ where we selectively attend to ‘proper objects of attention’. Minimally speaking, we have in CvD a car/driver and a cyclist/cycle in a road space where rules can be occasioned as relevant, amongst other things detailing who should be exactly where in the road. From this, there are two aspects of this tap on the car worth noting. First, it constitutes a minimal collision by Cyclist with the car, and second it is akin to what Burdelski and Cekaite (2020) call a ‘control touch’, that is, it constitutes a relationship in which one person exerts control over others, often being ‘laminated’ on talk. Both seem part of the reason for why Driver reacts so strongly and quickly. His direct point with the index finger targets Cyclist as the referent and also characterises him as some kind of wrongdoer (see Cooperrider, 2014), contrary to the effect of a control touch by Cyclist. In vernacular expression, this is also known as the ‘finger of blame’ (Givens, 2021).

Similar to the CvC episode analysed above, an emblematic gesture is then employed by Cyclist in response to Driver’s actions (pointing, shouting and a veer to the left). The silhouette (panel 6) shows Cyclist giving Driver a finger, as he says ‘no, no, fuck off’. In order to make this gesture, he briefly takes his right hand off the handle bar and turns his head right, but quickly returns to home position as he also speeds up. After 5 seconds (panel 7) we can hear the car revving from behind and Cyclist turns his head again to the right, takes his right hand from the handlebar and stretches it along his body in a protective manner. At this point, the car overtakes Cyclist, Driver shouts ‘pull over’ and stops on the left side of the road, in front of Cyclist (panel 9). However, Cyclist shows no intention of following Driver’s orders, but simply moves around Driver to the right and exclaims ‘you're on camera, so fuck off’, to which the Driver – already stopped – responds ‘fucking idiot’. Cyclist then cycles ahead.

In the final scene of the clip, which comes another 8 seconds later, Driver finally succeeds at making Cyclist pull over by aligning with him and decreasing the speed slowly while stopping accurately behind a parked car and cutting Cyclist off (see panel 1, Figure 6). Thus, with both vehicles stopped, the interaction becomes stationary at the 00:40 mark of the video clip. While stopping, Driver’s head is turned left to Cyclist (panel 2), repeating ‘you're a fucking idiot’, while Cyclist says, ‘you pulled out in front of me (.) it’s all on camera’ (panels 2, 3). While uttering ‘you’, Cyclist points with his right index finger to Driver, repeating Driver’s first declarative pointing at Cyclist. In response, Driver opens the right door and gets out of the vehicle, walking around its rear towards Cyclist. It can clearly be seen that his right fist is clenched, but as he walks around the car, he progressively loosens the grip (panels 7, 8). Meanwhile Cyclist, whilst finishing his utterance (‘it’s all on camera’) and seeing the Driver moving towards him, appropriates his bike in a defensive manner and probably (as we cannot see it just now) at this point grasps the D-lock (panel 9).

When Driver approaches Cyclist, who is holding the D-lock and exclaiming ‘get back!’, he keeps his right hand at his side and swings his left hand in a ‘put that thing away’ manner, which appears to be oriented to the D-lock in Cyclist's hand. This move is aligned with an agitated repetition of ‘you're a f:::fucking idiot’, with the ‘acme or thrust’ of the hand movement (Schegloff, 1984) timed precisely with the prosodically emphasised syllable of the expletive. ‘Get back!’ from Cyclist follows in response. Driver’s ruthless assessment of Cyclist is then repeated once again, this time with his body leaning forward and his face closer to the camera, no apparent hand-work involved. After a third ‘get back!’ from Cyclist, Driver returns to his car seat. While opening the door with his right hand, once again he repeats ‘you’re a fucking idiot’, precisely timed with a left index finger movement pointing towards Cyclist.

There is still nearly 40 seconds of the video remaining, during which significant things do happen, however, space constraints preclude further analysis of these and given our focus on interaction on-the-move, the stationary sequence is of lesser relevance, hence we move to a concluding discussion.

3. Discussion

In his explanation of the notion of social action, quite interestingly, Max Weber invites readers to imagine a situation not unlike the video recordings we have analysed above: ‘Not every type of contact of human beings has a social character,’ writes Weber, ‘this is rather confined to cases where the actor’s behavior is meaningfully oriented to that of others. For example, a mere collision of two cyclists may be compared to a natural event. On the other hand, their attempt to avoid hitting each other, or whatever insults, blows, or friendly discussion might follow the collision, would constitute “social action”’ (Weber, [1922] 1978, p. 23). In many ways, our analysis above has shown just how apposite the ‘on the other hand’ phrase is, for the conflictual actions on the move are absolutely saturated with see-sawing hand-work. The embodied practices to which we have turned our attention constitute highly varied and interconnected social actions, be it Beard’s insulting and dismissive ‘wanker’ gesture, his hand movements at the end in some attempt to make amends, or Cyclist’s tap on the car and the gesticulations it sets in train from Driver. Furthermore, both interactional sequences begin with a tacit response to a perceived digression from the normative order of traffic (cf. Lloyd, 2017). In the first case, the response is uttered – ‘come on mate’ by Beard; in the second case, the response is embodied – tapping on the mirror. In both cases, this initial response to perceived rule transgression is not explicitly formulating a rule. An explication comes only in the later reasoning and in both cases seems to be consequential in dismissing the emotionality of the encounter, and indeed the very reason for interacting as such.

The time and space limitations of interacting on-the-move cannot be forgotten, and perhaps it is because of such constraints that people may move to ‘semiotically loaded’ communication units such as emblematic gestures and expletives. There just might not be sufficient time for the sequential development of escalation as described by Hoey et al. (2021): a ‘normative ordering whereby participants first manage interactional difficulties through relatively tacit procedures, and then given the failure of those, through more explicit and escalated ones’. When an everyday trouble appears on the road – why isn’t that guy moving?; what on earth is he complaining about? – perhaps launching a confrontational action is a readily understandable ‘natural reaction’.

In relation to Katz’s useful work with which we began, we agree that actions on the road are enigmatic and can change rapidly, but now having seen some of the embodied and gestural aspects of two conflictual incidents, we need to be cautious in adopting Katz’s characterisation of them as idiosyncratic. On one hand, mirroring Streeck’s observation regarding the organisation of attention to gesture, ‘we assume that we know what the gestures refer to (and indeed, in most cases there is little doubt about this)’ (2009, p. 87). On the other hand, he also says in agreement with Katz, ‘pragmatic gestures are an unruly bunch: speakers show all manner of idiosyncrasies in making them’ (2009, p. 181). This just serves to emphasise the need for detailed empirical scrutiny of instances, like that we have begun to provide above.

The six graphic transcripts we have used above clearly cannot cover all the detail of the studied interaction sequences, and of course there are limitations in what the data enable us to pinpoint; however, in summary we would like to emphasise four key points.

1. The two cyclists in CvC are strangers to each other, in motion in a mundane space with spatial limitations as to where a cyclist can safely proceed. This is part of the initial complaint: In Beard’s view, Camera is unnecessarily holding him up, but in response to Camera’s justification – safety – for his action, Beard quickly moves into the abusive counter, ‘if you’re scared you shouldn’t be on the fucking road’. Similarly, in CvD, Cyclist and Driver are strangers to each other, but in this case even though Cyclist is more at risk from a collision with a car, it is Driver who takes umbrage at the ‘collision’ – the tap – with his car. Thereafter, Driver’s abuse of Cyclist is predominantly a combination of declarative pointing coupled with the exclamation ‘you’re a fucking idiot’. This perhaps could be seen as a minor-level conflict; however, we should not overlook the more concerning actions he employs in using his vehicle to stop Cyclist, with what looks to be an intention to engage in assault.

2. In CvC, the first abuse is made while riding alongside at a faster pace, and as Camera immediately counters, Beard has moved several metres ahead, which is where a very significant gestural moment occurs. Beard makes a hand motion of ‘wanker’ as he rides off in a ‘departure sequence’ (LeBaron & Jones, 2002). For the two cyclists, in the real time of the situation, this gesture could have been the final moment of their interaction - an ‘improper closing’ (Frank, 1982). In contrast, in CvD the first significant hand action is the tap by Cyclist on the wing mirror, and this immediately sets Driver into a strong reaction. This includes both the accusatory pointing coupled with the ‘fucking idiot’ characterisation, and the veering leftwards, the latter being a dangerous deployment of his vehicle. Katz’s work offers a neat characterisation of these actions, as he emphasises that drivers, in order to drive, are embodied by the car in a tight human-machine assemblage (1999, p. 32). The driver has a range of connections to the car and ways of sensing it (cf. Salvadori & Gobo, 2021), which helps us understand why strong reactions can happen almost instantaneously. To extend Katz, this can apply whether it is a car or cyclist that ‘cut offs’ the driver. For,

What all experiences have in common is a kind of amputation, a loss of a previously engaged, tacit use of the car and a loss of the transcendent body that the driver, in the process of driving, had been taking for granted as naturally available. [Thus, we need to understand] how becoming angry is a practical project in which the driver attempts to regain a taken-for-granted intertwining with the environment. (Katz, 1999, p. 33, emphasis added)

From the CvD case we may add to Katz’s emphasis on what is taken for granted, drivers’ own self-characterisation of their driving. Later, when Driver has Cyclist stopped, he is adamant that he was driving correctly – ‘I kept right over to the right hand side’ – but this verbal account is reflexively tied to his earlier actions of abusing Cyclist and trying to get him to stop.

3. In CvC, Camera does catch up with Beard after the significant ‘wanker’ gesture, but there is no response from the latter, thus the traffic light stop a few minutes later is very significant, as now both cyclists are stopped. Additionally, visible in the background it is clear there are also other cyclists ‘overhearing’ their interaction. As noted, with Beard stationary beside him, thus available for interaction, Camera is doing much finger-pointing. We think this is key to the way in which his emotional state is ramped-up into ‘interactively generated anger’ (Robillard, 1996, p. 27). In regard to this, Streeck importantly argues that, ‘from the gesturer-speaker’s point of view, pragmatic gestures are a feedback mechanism that in effect tells the speaker what he or she is up to: gestures provide experiential frames … in terms of which our own communicative actions are tacitly made meaningful to us’ (2009, pp. 201-202). Camera does not get close to a physical self-fuelled spiral of rage (Scheff & Retzinger, 1991), nevertheless his seeing of his own gesturing work reaches its apex in the exclamation ‘shut ya fucking mouth’, simultaneous with an emblematic gesture (that we cannot see in full). The presence of both the ‘overhearers’ and the visibility of Camera’s growing emotional state may partly explain why Beard does not engage, rather he twice utters ‘fuck off’ while embodying an orientation to riding, that is, being prepared to ride off from the interaction with Camera. Similarly, in CvD, a passing-by jogger speaks to the dyad just in the middle of their heated argument, and Driver’s response to the passer-by is markedly calmer in comparison to his previous expletive-laden exclamations. The fact that both CvC and CvD happen in public spaces with actual or potential presence of other non-acquainted people could be consequential for the development of violent actions, which are projected (especially in CvD) but never really occur. We have spent less time describing CvD, but in relation to a characterisation of it as being closer to physical violence, it has to be emphasised that Cyclist’s defensive use of the D-lock seems highly effective. This may also help explain Driver’s recourse thereafter to expletive-laden abuse – ‘you’re a fucking idiot’ – which, as Collins (2008) has argued, is a kind of ‘bluster’ that generally does not escalate into actual physical fighting.

4. In CvC, shortly after Camera moves off ahead from the traffic lights, the two conflicting cyclists ride alongside each other (the only time they have been in a ‘mobile formation’; see McIlvenny, 2015), and now having just told Camera to ‘fuck off’ twice Beard engages in some version of remedial action. This does not take the form of an apology, rather he tries to justify his initial complaint. This is heavily replete with gesticulation, as here Beard is riding with both hands off his handlebars. This both enables his pragmatic gesturing, but somewhat speculatively we suggest it may have the meta-pragmatic function of showing that he is a very skilled (and risk-taking) cyclist; in short, he is speaking to Camera from the position of being the ‘better cyclist’. Not surprisingly then, Beard and Camera’s interaction ends with a separation of ways without agreement or appeasement, and indeed without a proper closing – similarly to CvD in this respect. Strangers in the initial moments, they part company as strangers.

4. Conclusion

In a well-known passage in Phenomenology of Perception, Merleau-Ponty provides an illuminating and fundamental insight into the relation of anger to its embodied aspects:

Faced with an angry or threatening gesture, I have no need, in order to understand it, to recall the feelings which I myself experienced when I used these gestures on my own account. I know very little, from inside, of the mime of anger so that a decisive factor is missing for any association by resemblance or reasoning by analogy, and what is more, I do not see anger or a threatening attitude as a psychic fact hidden behind the gesture, I read anger in it. The gesture does not make me think of anger, it is anger itself. However, the meaning of the gesture is not perceived as the colour of the carpet, for example, is perceived. […]. The sense of the gestures is not given, but understood, that is, recaptured by an act on the spectator’s part. The whole difficulty is to conceive this act clearly without confusing it with a cognitive operation. […]. Here we must rehabilitate the experience of others which has been distorted by intellectualist analyses, as we shall have to rehabilitate the perceptual experience of the thing. (Merleau-Ponty, [1945] 2002, pp. 214-215, original emphasis)

This crucial early attempt to think about emotion and anger in a non-cognitivist manner is now increasingly important for scholars of social interaction (cf. Peräkylä & Sorjonen, 2012). Katz, for example, has neatly emphasised how emotions are not in tension with thought, reason or self-examination, “[j]ust the reverse: emotions are lived as metamorphoses toward thought, as movements from an unself-conscious being-in-the-world to relatively more self-reflective postures” (2002, p. 261). And there is Streeck’s powerful work on gestures (2009, 2017, 2021), particularly his rejection of the ‘instrumental’ model of embodied action; that is, the notion that the body is under the mind’s control, for ‘as everyone knows, feet can tap, knees shake, and hands fidget without anyone making them do so’ (Streeck, 2020).

As we have shown, the work of hands seems to be intricately tied to emotionally charged situations, and hands do what they do as constitutive of anger itself. Clearly, there is subtle coordination work engaged in by cyclists and drivers as they move together in specific road spaces, hence they must constantly work at reciprocal adjustment. Decisions about what to do and where to move are compressed, and sometimes must be co-extensive with its being done. The visibility and tactility of hand-work helps provide mechanisms of expectation and feedback, forward-projection and retroaction, which, as we have clearly seen, can lead to open conflict on the boundary of physical violence. The video data we have utilised above does not provide a perfect entry point into the study of these processes; nevertheless, careful attention to its detail allows the discovery of just how finely ordered the hand-ling of road rage can be.

The majority of the literature on road rage centres around automotive-based conflict, but clearly it can happen in almost any kind of action on the move. Whilst looking for similarities across different forms of mobility, we also need to remain attentive to specific differences. Cyclists are on a bike, not enclosed in it, therefore they are both more vulnerable in a collision – the ‘big truck’, the ‘sports car’ – and afforded greater visibility and hearability resources. For example, the ‘wanker’ gesture is easily made by Beard, whereas in a vehicle a driver would often have to open a window, then extend an arm out to make the action, all at a moment when the targeted other is close enough to see, all whilst steering the vehicle. Moreover, we need to remain open to significant differences in the form of hand-work. As Cuffari and Streeck neatly summarise, ‘[h]and gestures both fit the world at hand and form it. The fitting has to do with what from the environment gestures appropriate. The forming has to do with what gestures disclose – that is, what they reveal, forefront, show in a new light, and create’ (2017, p. 176, original emphasis).

Without dogmatically extolling the virtues of an ethnomethodological approach, here we have tried to work from the detail-up. Rather than theorise or conceptualise ‘road rage’, or gesture for that matter, we have carefully looked at the details of two short publicly available videos that do come with ‘road rage’ in their titles. We invite any reader to view them and feel confident that such a viewing will acknowledge the significance of the hand-work in the interactional to and fro on the boundary of physical violence. It is not all that happens, but not noticing it would be to miss a lot. Such attention to hand-work could be usefully incorporated in further interactionist studies of road rage, including an extension to different modes of being mobile in traffic spaces.

Acknowledgements

Both authors thank the anonymous reviewers and the journal editors who made useful and valuable comments on the first draft of this article. We are grateful to members of the Interaction and Social Practices Research Group (GRIPS) at the University of Fribourg, namely Esther González-Martínez, Bastien Taverney, and Sylvia Trieu, for their inspiring suggestions and comments. Jakub Mlynář would also like to thank the Charles University Research Centre No. 9 (UNCE VITRI), which supported his work on this article.

^{1 In the discussion below, when it is appropriate, we utilise specific technical terms from gesture scholarship, however, the point needs to be emphasised that our goal is not thereby to enter into the ‘terminological tangle’ (McNeill, 2005, p. 21) that can follow classificatory moves in gesture studies. We often rely on lay terms, and in general endorse Kendon’s neat definitional move regarding gesture: “Utterance, thus is a polymodal activity. The movement of the hands and arms in particular, generally referred to as ‘gestures’, are usually regarded as expressing something that is part of the meaning the speaker is conveying” (2011, p. xv, original emphasis). This is offered in a foreword to Calbris (2011), a work offering a comprehensive classificatory approach to the semantic analysis of co-speech gesture (also see Clarke et al., 2021 for a ‘protocol for gesture analysis’). It can also be emphasised that one reason for our eschewal of technical terms is the difficulty of pinpointing the precise role of gestures in interaction which occurs while interactants are mobile, sometimes at significant speed. Also see Clift (2020) for an analysis of the ‘Palm Up’ gesture, where the argument is made that whereas gesture scholars have attempted finer delineations of this gesture, a conversation analytic stance can eschew such terminological branching to focus on the interactional sequences within which gestures appear.↩}