Emergency or Not? Dealing with Borderline Cases in Emergency Police Calls

ABSTRACT

We examine occasions when callers phone emergency services yet preface their reason for calling as “not an emergency.” Data are phone calls to U.S. (911) and U.K. (999) emergency lines and U.K. (101) nonemergency police lines. Data were transcribed using Jefferson conventions and analyzed using conversation analysis. The “not an emergency” formulation is recurrently used to mark a shaky or borderline fit between the caller’s situation and the emergency category presumed by the dedicated phoneline. Typically, “not an emergency” formulations prefaced descriptions of a possible emergency in which the caller balances the justification for the call on the boundary of what counts as an emergency. Recurrent concerns for callers using “not an emergency” are to manage preemptive calls about impending potential emergency, and to disclaim responsibility for the decision to call an emergency service. Call takers offer callers latitude to present a complicated description of their circumstances instead of swiftly sanctioning them for an inappropriate call. Our article contributes to work on how the boundaries between categories are constructed and negotiated in interaction. Data are in British and American English.

In emergency calls, callers’ first turns are critical for establishing the reason for calling and setting the trajectory for collecting pertinent information for dispatching first responders. On some occasions, however, callers produce a seemingly self-contradictory action by prefacing their request with, “It’s not an emergency,” thereby beginning their call by stating they are not experiencing the required categorical circumstances that entitles them to request emergency assistance. Here we analyze such instances for what they can reveal about the boundaries of categorical attributions in interaction.

Emergency call openings

Emergency calls to public service telephone lines, such as 911 in the United States and 999 in the United Kingdom, are social occasions requiring the participants to orient to, and make available through their conduct, a specific institutional context (Drew & Heritage, 1992; Kevoe-Feldman, 2019; Schegloff, 1987; Zimmerman, 1984, 1992). Members’ displayed orientations to the routine ordering of activity provide evidence for members’ mutual orientations for enacting context. For example, specialized openings by call takers orient to an institutional self-identification (e.g., “911, what’s your emergency”) rather than a “recognition-oriented response to the telephone summons” (Whalen & Zimmerman, 1987, p. 175). The opening sequence establishes a relevant context for callers to report emergencies (Zimmerman, 1992) and places a constrained focus on the caller’s first turn as the reason for calling emergency services (Whalen & Zimmerman, 1987). As Cromdal et al. (2012, p. 375) put it, “[T]he onus is on the caller to immediately supply a relevant reason for initiating the contact.”

Whalen et al. (1988) provided an account for how interactional asymmetries become salient when parties’ orientations and expectations emerge in the emergency call encounter. They explained that during call openings participants align their identities as service provider/service seeker. The mutual identification of emergency service provider and emergency service seeker sets up categorical circumstances that (a) make the caller’s first turn hearable as a request for urgent help (over and above other potentially relevant actions such as troubletellings, Jefferson & Lee, 1981), and (b) position the call taker as responsible for deciding what (if any) assistance will be provided by the organization.

Most emergency organizations described in the literature specify a particular call opening formulation for their call takers (for an overview, see Cromdal et al., 2012). The same is true for the different corpora we analyze in this article. There is some evidence that the format of the call takers’ opening influences the trajectory of the call. Cromdal et al. (2012) found that, compared to identification only, call taker opening turns that ended with a direct query about the incident (e.g., “112, what has occurred”) more often solicited descriptions of the trouble instead of requests for help and facilitated a smoother transition to the interrogative phase of the call. However, given the institutionally mandated consistency of the call taker’s opening turn, outside of comparative studies like Cromdal et al. (2012), it cannot fully account for how or when callers choose to preface their reason for calling with, “It’s not an emergency.”

Callers reasons for calling

There is considerable variety in how callers choose to formulate their reason for calling the emergency services at multiple levels of granularity. Kent and Antaki (2020) reported distributions of requesting, reporting, and descriptive first formulations during police calls that broadly concurred with Whalen and Zimmerman (1990) and Drew and Walker’s (2010) characterizations of their US and UK emergency call datasets. They identified descriptive formulations as the most common (49%), following by requests without contextual descriptions (28%), and descriptions containing explicit requests (18%), with reports and other formats being least common (5%).

When examined in their interactional context, the turn design variations of callers’ reasons for calling initiate different contingencies that call takers must manage. Whalen and Zimmerman (1990) found that participants orient to different vulnerabilities associated with their descriptions about emergency events. They systematically documented how participants manage these vulnerabilities through an orientation to “displays of a practical epistemology—just how, on this occasion, one has come to know about this particular event” (Whalen & Zimmerman, 1990, p. 466).

Similarly, G. Raymond and D. H. Zimmerman’s (2007) work demonstrated how callers’ delaying their request for emergency service set up interactional contingencies that alter the routine progression for the request and the ultimate fulfillment of emergency service. During community emergencies such as fires, flooding, and blackouts, callers may presume the emergency was already reported. Thus, “there is a tension between the caller’s right to report the event and his or her obligation, as a competent member of society, not to tell a recipient something they already know” (G. Raymond & D. H. Zimmerman, 2007, p. 42). To manage this tension, callers to emergency service will begin their reason for calling with a pre-sequence designed to elicit (dis)confirmation from the call taker to clear the way for their report about the emergency (e.g., fires).

Callers routinely upgrade or downgrade the urgency of their reason for calling through turn designs that embody different levels of entitlement for service. Callers’ orientation to low/high entitlement for emergency services influences the call handling procedures. Larsen (2013) demonstrated how call takers are sensitive to the various linguistic formats callers use to design their requests. She observed that when callers design a request with a strong sense of entitlement by using a modal verb such as “jeg/vi skal–I/we must” (p. 214) and include relevant institutional information as part of their reason for calling, call takers match the caller’s display of entitlement by asking questions required for granting their request. Moreover, she also found different cases demonstrating that when callers mitigate their request with downgraded claims of entitlement, dispatchers forward the progressivity of the call but ask incident-related questions as opposed to matters directly related to the act of dispatching emergency services.

Relatedly, Kent and Antaki (2020) showed how callers design their request for police assistance, shape the trajectory of the call taker’s first question of the interrogative series, and ultimately project the call’s outcome. For example, when callers initiate their request with a high entitlement request format (e.g., “I need you to send the police”), call takers’ first questions are typically about locations and conveys a dispatch-implicative move that projects fulfillment of the request. However, when callers design their reason for calling with low entitlement or their report is not straightforward (K. Tracy & R. R. Agne, 2002), call takers typically begin the interrogative series with more probing questions that work on unpacking the policeable nature of the request, thereby projecting a possible outcome of not granting the request. Fele (2014) found that when callers design their request with high entitlement and it then emerges during in the interrogative series that they have insufficient knowledge about their emergency, they run the risk of the call takers rejecting the request for assistance. However, if the caller prefaces their lack of knowledge about the event before call taker’s questioning, then the call takers will proceed with the line of questioning formulating inquiries within the caller’s knowledge domain.

Our analysis explores how the practice of prefacing the reason for calling with “it’s not an emergency” helps callers manage a specific kind of practical problem. First, we show how this practice makes visible the categorical boundary of what constitutes an “emergency,” and from whose perspective, by downgrading the caller’s entitlement for service. We show how callers use this practice when reporting incipient emergencies, which helps manage the problem of calling for help before something has happened. Finally, we show how participants negotiate the rights and responsibilities of categorizing a situation as an emergency, and how callers enlist the assistance of the call taker to make that final determination.

Categorization

Extensive work by interaction scholars has explored categorical identity membership and its use in interaction (e.g., Fitzgerald & Housley, 2015; Robles, 2022; Stokoe, 2003; Stokoe & Edwards, 2009; Tennent, 2021). This line of work has examined what Sacks initially termed “category-bound” activities “taken by members to be done by some particular or several particular categories of members” (Sacks, 1992, p. 249). Such work has revealed the highly attuned and delicate nature of how speakers manage and draw on categorical identities when performing a wide range of actions in social interaction (C. Raymond, 2019). For example, identity categories make relevant category-bound rights to perform certain activities, such as assessing grandchildren’s behavior (G. Raymond & J. Heritage, 2006).

Given the importance of categories for the organization of rights and memberships in interaction, how categorical boundaries are constructed and oriented to in talk is pertinent. Rossi and Stivers (2021) explored how social memberships are tied to category-sensitive actions. They acknowledged that for most of the time categorical boundaries might remain invisible and are only topicalized when their boundaries are approached, exposed, or transgressed in some way. Our choice to build our analysis around the target phrase “not an emergency’ facilitates the exposition of how the categorical boundary of an emergency is constructed and negotiated at the very moment in social life when it is at its most consequential—when deciding if formal emergency services will provide assistance to citizens requesting it.

Far less attention has been paid by interaction researchers to the interactional organization of categories for action, activities, or objects less directly tied to a speaker’s identity displays. What is at stake in our data is whether a situation (as described by the caller) meets the categorical threshold for an emergency (as determined by the call taker on behalf of the emergency services). In this respect, our work has some parallels with how diagnostic labels are attributed during healthcare interactions. Turowetz and Maynard (2016) described a set of practices associated with what they termed “category attribution” when diagnosing a child with autism. This is a process through which the narrative about a child’s experiences is juxtaposed with claims about members of the clinically relevant category, with the implication that the child does or does not belong to that category. They suggested that category attributions are the interactional means through which idiosyncratic cases are fitted to general diagnostic categories. For emergency callers and call takers, how members establish an emergency categorical attribution during an emergency call is a central component of the interaction. The extreme time sensitivity of an emergency call affords an accelerated interactional context to explore how institutional (or diagnostic) categorization is organized.

Our analysis exposes the interactional work required at the boundaries of multiple possible categorical descriptions of a situation that make relevant different institutional treatments (e.g., emergency response/nonemergency police action/no action/sanction for vexatious contact). Here we examine the interactional implications of when emergency callers preemptively and explicitly topicalize the categorical fit of their circumstances by prefacing their reason for calling with, “It’s not an emergency.”

Interactional asymmetries in emergency calls

Drew and Heritage (1992) argued that institutional interactions are inherently asymmetrical in organizing and managing turns and knowledge (Drew & Heritage, 1992; Maynard, 1991). Individuals who seek out professional assistance generally treat the professional as having more knowledge or expertise about the matter than a layperson. K. Tracy (1997) explicated how asymmetries in emergency calls can lead to interactional trouble when there is a mismatch between what the caller wants and the organizational constraints a call taker must follow. At times, this can mirror the interaction asynchronicity described by Jefferson and Lee (1981) for troubles-tellings and service encounters. Callers may not understand why a call taker must ask a series of questions for information gathering when, from their perspective, they have provided the call taker with all the relevant information to proceed (S. Tracy, 2002, 1997; see Heritage & Clayman, 2010). Callers can treat a line of questioning as threatening their epistemic position relative to the emergency, thereby threatening their status as a trustworthy source (Garcia & Parmer, 1999; S. Tracy, 2002).

Our analysis broadens our understanding of variations in callers’ requests for emergency service that alter the routine progression of emergency call-handling procedures by exploring the interactional implications of explicitly topicalizing categorical descriptions of emergency vs. nonemergency situations at the outset of a call.

Data and methods

Our core data collection draws on a corpus of 470 telephone calls made to a regional U.K. police call center answering emergency (999) and nonemergency (101) phone lines, and 1,432 telephone calls made to a wireless emergency communication center in the United States (see Kevoe-Feldman & Pomerantz, 2018 for a description of the U.S. call center). The dedicated U.S. wireless line receives incoming calls from wireless phones only (not landlines) and transfers callers to the local responding emergency service as appropriate. In the United Kingdom, call takers can initiate police assistance (emergency response or routine follow-up) equally easily from both 999 and 101 phone lines and have equivalent access to patrol dispatch.

Each author’s university review board provided ethical approval for the separate data collection sites, and permission was gained from the participating emergency call centers. All identifying information, participants’ names, institutional affiliation, and location information such as address and telephone numbers have been redacted to protect the participants’ identities.

Analytic approach

Calls were transcribed using the Jefferson conventions (Hepburn & Bolden, 2017) and analyzed using conversation analysis (Schegloff, 2007; Sidnell & Stivers, 2012). We identified cases in which the caller’s first formulation of their reason for calling includes the phrase “not an emergency.” This collection also included variations with simple modifiers (e.g., not really an emergency, not a serious emergency). We did not restrict our collection based on the position of “not an emergency” within the caller’s first formulation of the reason for calling. However, it almost exclusively appeared as an initial turn feature. We also gathered a contrasting collection of related cases to illuminate relevant features of our developing analysis (Kevoe-Feldman & Iversen, 2022). For example, 101 (nonemergency) UK police calls enable us to explore uses of the target phrase during institutionally comparable calls that lack the contextual and structural presumption of an emergency inherent to 999/911 calls. The more ad hoc contrast collection included these:

• 101 (U.K. police nonemergency) calls in which the caller used “not an emergency”;

• other items within the first formulation that oriented to the suitability of the call to an emergency line (e.g., “I wasn’t sure if I should phone 999 or 101 with this”);

• calls in which “not an emergency” occurred after the first formulation had been completed; and

• examples of very similar incidents from different callers in which not all chose to include “not an emergency” (e.g., power lines down across a road).

Of the U.K. calls, seven cases feature our main phenomenon, with 38 offering contrast cases or other relevant features that supported our developing analysis. For the U.S. data, 25 cases feature our main phenomenon, with 23 supporting cases. Although “not an emergency” is a rare feature of our datasets, our analysis revealed that it represents a robust phenomenon in turn design and action orientation.

Analysis

As previous researchers have demonstrated, there is considerable contextual and structural pressure for callers to articulate a recognizable emergency as their reason for calling (Cromdal et al., 2012; Whalen & Zimmerman, 1987; Zimmerman, 1992). Callers can be (and routinely are) held accountable for failing to promptly provide a categorically fitted reason for calling. For example, in Extract 1 the caller offers to provide a crime number (for an ongoing incident) “first”—before providing the reason for calling (lines 3–4). Instead of accepting the offer, the call taker seeks confirmation that their call is about an emergency (line 6) and justifies that prerequisite by reminding the caller that “yur calling on a nine ni:ne line” (lines 6–7).

Extract 1. P208 999 Update on death threats

In response, the caller does more than just confirm the call is about an emergency (“yeah i- yeah” line 8); they index recent contact with an officer about a life-threatening matter about which they have been given third-party permission to call the emergency line (lines 10–14). When challenged, we see that they immediately justify the categorical fit between their situation and the emergency phone line. They do not persevere with providing the crime number first.

Given the structural and contextual pressure on callers to produce a categorically fitted reason for calling the emergency services, our analysis explores what affordances the “not an emergency” preface offers callers that warrant its insertion into the first formulation of their reason for calling. Our analysis focuses on instances in which callers delay their request for emergency services by prefacing their reason for calling with “not an emergency,” as in Extract 2.

Extract 2. 911 Dangerous driver

In Extract 2 (and throughout our core collection), the first thing the caller does is explicitly categorize their reason for calling as not an emergency—that is, not fitting the constraints of the interaction that are both preestablished (by choosing to phone a dedicated emergency line) and locally established through the call taker’s specialized opening (Zimmerman, 1992). In Extract 2, the caller’s response that “It’s not an emergency” (line 5) displaces and delays the start of their report of a dangerous driver. Our analysis outlines how callers use “not an emergency” to (a) make visible their orientation toward actions that may fall outside the categorical boundary of “emergency,” (b) manage the in situ progression of an emerging situation, and (c) disclaim responsibility for evaluating a situation as warranting emergency assistance.

Emergency category boundary orientations

Through the “not an emergency” formulation, callers explicitly orient to the boundary of what constitutes a legitimate request for “emergency” service and preemptively position their reason for calling as a problematic fit with how the institution categorizes emergencies. We observe how a caller’s first turn swings back and forth between claiming and disclaiming an entitlement to seek emergency assistance. This oscillation between opposing stances toward the legitimacy of their call produces a delicately counterbalanced turn, as shown in Extract 3.

Extract 3. 911 Debris on road

In the first place, the caller chose to call a dedicated emergency line, thus claiming the putative right to request urgent police assistance. Then, at their first opportunity to present a reason for expecting emergency service, they insert “Yeah. It’s not rea:lly an emergency” in line 4, which positions the situation as not a true (“real”) fit for emergency categorization. This weakens their entitlement to request immediate assistance. In the next component, the caller describes and assesses a situation explicitly marked with emergency implicative features (“debris right in thee uh travel lane,” and assessing it as “It’s kinda dangerous,” lines 5–6), thus building a case for how this “not real” emergency might qualify as one. The description counterbalances the “not an emergency” preface to keep the caller’s overall positioning of their circumstances as deliberately teetering on the edge of what might be categorized as an emergency. In this case, the caller then makes explicit their own oscillation about whether the circumstances could be categorized as an emergency by stating their uncertainty about “who to call” (line 9).

Extracts 4a and 4b further exemplify the boundary oscillation produced in turns using the “not an emergency” formulation.

Extract 4a. 911 Locked out

Here the caller orients to the not “overly huge” (line 3) nature of the emergency, marking the possibility that his circumstances are not sufficiently dire for his emergency status to be beyond doubt and positioning himself as knowingly calling about something potentially ineligible for emergency assistance. The second part of the turn offers an ambiguous location, questioning whether the emergency line he called could assist, instead of a description containing emergency-implicative information (lines 3–5). In response, the call taker directly queries, “what are you callin’ for?” (line 6). Positioned as a post-first insertion sequence (Schegloff, 2007) prior to confirming (or disputing) that the caller has phoned the right place, the call taker’s open-ended information solicit (line 6) makes his response contingent on the caller’s ability to describe a situation that would warrant an emergency response (c.f., Kent & Antaki, 2020). Thus, seeking to clarify the categorical fit of the reason for calling with the remit of the emergency phone line.

The caller then describes a situation in which he is a long way from home (lines 8–9) and has lost access to both his vehicle and his possessions (line 10), rendering him effectively stranded. He upgrades the seriousness of his circumstances by reporting that his children and family are with him (lines 10–11). Through the contrastive “but” (line 3) turn design, these features are presented as evidence of an emergency. Extract 4a exemplifies how callers first claim entitlement for emergency services by phoning an emergency line yet preface their reason for calling by disclaiming such entitlement with, “It’s not an emergency.” By using a descriptive rather than a directive formulation to minimize their display of entitlement for police assistance (Larsen, 2013), callers provide information for the call taker’s professional assessment of whether the situation warrants service. Through this design, callers mark the boundary of what constitutes a known-in-common emergency and embody an awareness that their circumstances might fall close to that boundary. In so doing, they avoid overasserting their entitlement to expect emergency assistance while not relinquishing the possibility.

This case is illuminative because the call is transferred to a local police operator. Locally subsequent versions of utterances are analytically valuable for what they retain, drop, or gain compared to the original (Schegloff, 2004). When callers phone 911 from their wireless device, the emergency will ring to a dedicated wireless call center. That call taker will gather pertinent information and transfer the caller to the responding town. Extract 4a was the caller’s first call to the wireless call center. Extract 4b is the same caller but now transferred to a local call center, where the caller needs to report the emergency for a second recipient (who was not privy to the first telling).

For his second version, the caller omits the “not an emergency” preface and delivers an upgraded description of his circumstances with increased markers of the severity and urgency of his situation. The caller has already passed the gatekeeping of the first operator without being dismissed for making a frivolous call. The ratification ostensibly offered by that call taker transferring the call might account for why the same situation is now not constructed as a borderline emergency (although it is still not expressed as a strongly entitled request such as, “I need the police right now”).

Extract 4b. 911 Locked out; transferred call

¹

This time, the “seven-year-old kid” (line 4) is foregrounded as the first piece of information and expanded with “a family” of unspecified size. This formulation highlights the vulnerable nature of the people involved in the emergency—to whom the police may owe a higher duty of care. Then the caller explains that the keys are locked in the “trunk of the car” (line 7) and highlights he has exhausted other options for assistance (“these people won’t help me whatsoever,” line 8; “we don’t have triple A,” line 9; “nothin,” line 9) positioning himself as a responsible citizen who tried to solve the problem before calling the police (c.f., Heritage & Robinson, 2006 on doctorability). In this transferred call, the caller designs his request to maximize the justification for calling emergency service for a family stranded in an otherwise safe hotel car park.

Despite the caller’s efforts to foreground emergency-implicative features, there is no reference to current or imminent danger or harm. Objectively, one might argue the family is inconvenienced rather than in peril, although there is the implicit notion that harm might come to them if they cannot access their vehicle. These implications contribute to the sense of this being a marginal case for the category of a police emergency. The call taker’s ultimate disposition of the call (lines 10–11) supports this analysis. Instead of offering police assistance, the call taker explains that the fire department is the appropriate response agency and transfers the call (again).

Incipient emergencies

When considering the boundaries of what constitutes an emergency, the time-sensitive nature of such situations is a key consideration. A recurrent concern for callers using the “not an emergency” formulation was to preempt an impending, but not yet realized, emergency. A caller’s “not an emergency” often prefaced a report of something that could turn into a serious or urgent situation imminently, such as road blockages that could cause collisions (e.g., Extract 3). Potential traffic collisions were common in our collection but by no means the only incipient emergencies that callers reported. Extracts 5a and 5b are from a preemptive call in which urgent police intervention would be required should the situation develop as feared. The caller is concerned that her bus driver will not stop for her to disembark.

Extract 5a. 911 Bus not stopping

The caller signals a strong orientation to a potentially nonfitted reason for calling emergency services by explicitly stating a lack of knowledge about whether it is an emergency or not (line 4). Her epistemic uncertainty about the limits to the emergency category draw attention to the boundary (see also Fele (2014) on how the caller and call taker must manager her overall uncertainty about whether she is in danger or not). The following description of the emergency “I’m on the bu:s” (line 5) does not offer a counterbalance to the preemptive flag of a potentially inappropriate emergency call: It contains no explicit or objective markers of an emergency. This might explain the call taker’s noncommittal sounding “yeah” (line 6). First formulations that do not contain any information that might help the call taker make a judgment about institutional fittedness are often responded to minimally and with continuers to treat the formulation as inadequate or incomplete (Kent & Antaki, 2020). Only after the caller constructed the bus driver as actively preventing her from leaving the bus (line 9), making relevant police concerns around unlawful detention or abduction, does the call taker begin to engage with the situation as a candidate emergency (lines 10–12).

Omitted are 21 lines of transcript in which the call taker tries to solicit the caller’s location four times with increasing urgency. At the same time, the caller provides a narrative account of her interactions with the bus driver. While the call is active, the caller’s situation changes, and the incipient emergency is averted (Extract 5b).

Extract 5b. 911 Bus not stopping

On lines 34–39 the caller reports that she is no longer in danger because the bus is now preparing to stop where she expected it to. The change of status from an incipient emergency abduction to a routine bus journey is marked by apologies from the caller for ultimately wasting police time (lines 34 and 39), but also is attributed to the driver’s awareness that she was talking to the police (lines 36–37), thus reiterating the caller’s stance that the threat was real, and the preemptive call was warranted. Even in its resolution, the call remains balanced on the edge of the emergency category.

Relatively short time increments make the difference between an emergency and not. By foregrounding the boundary of the emergency category, the “not an emergency” formulation orients both the importance of promptly phoning to access urgent assistance and the risk that the call may prove premature and unwarranted as time passes. This is one of the key boundary balancing acts performed by “not an emergency” formulations.

Extract 6 is from a 101 (U.K. nonemergency) police call. Without the pre-call orientation of the call to a dedicated emergency line, the “not an emergency” preface seems particularly redundant as it does not oscillate over the emergency categorical boundary. Unlike previous examples, this caller is “totally” convinced that her call is not an emergency (line 3). This epistemic certainty is born out as it emerges that the “not an emergency” formulation prefaces a request for police assistance to prevent a relatively trivial problem from occurring. In this case, the boundary being oriented to is one of police relevance rather than the higher threshold for emergency response.

Extract 6. 101 Bike lock

The time frame within which the caller’s stuck bike might get stolen (tomorrow) is far longer than the time frame in which road debris might cause a crash (Extract 3) or a bus might be hijacked (Extract 5). However, despite the lack of urgency, the same considerations are in evidence: Police assistance is being sought to preempt a situation that would relevantly involve the police (reporting a stolen bike).

Note that the caller is not reprimanded for a frivolous call. In fact, the call taker uses a “no fault” account when explaining the decision not to send assistance (Heritage, 1988). She is gently led to conclude that police assistance is not proportional to her problem (only armed response officers have bolt cutters, and they are not available to unlock citizens’ bikes; lines 20–25). Across our data, it appears that a preliminary admission that the reason for calling is an ambiguous emergency might help to inoculate callers against negative judgment for their decision to call, as long as their subsequent reason for calling contains at least some evidence that counterbalances their early disclaimer.

Entitlement to classify something as an emergency

There is a structural presumption when calling dedicated emergency phonelines that the caller has preevaluated theircircumstances as warranting the category of emergency. We suggest that when callers use “not an emergency” prefaces, they position themselves as lacking the entitlement to classify their situation as an emergency, despite their choice to use an emergency phone line. Extract 7 exemplifies this through a self-repair exposing a shift in the caller’s stance toward her entitlement to assert that her situation is an emergency.

Extract 7. P256 999 Stolen purse

The caller’s news receipt (Heritage, 1998) at the start of line 3 (“Ooh”) appears to register the call taker’s use of the word “emergency” (line 1) as a category the caller had not previously applied to her situation, despite calling an emergency phone line. She begins to say “its not a- … ” but then breaks off and transforms her utterance into “it is an emeractu’ly” [emergency actually] (lines 3–4). This repair makes visible her in situ evaluation that she is “actually” warranted in claiming that the emergency category applies to having one’s purse stolen. Note before ratifying her claim, the call taker first checks the timeframe since the theft: Pick pocketing only warrants an emergency response if there is a reasonable chance of catching the thief before they leave the area. The caller’s uncertainty on line 3 orients to the ambiguities about her circumstances’ categorization as an emergency. The “not an emergency” preface was almost used to straddle the boundary line before she ultimately chose to make a stronger claim.

“Not an emergency” formulations orient to how callers and call takers negotiate the rights and responsibilities to categorize situations as emergencies and seek/provide official assistance. Phoning an emergency line and preemptively declaring that your call is not an emergency generates an ambiguity that places the boundaries of what constitutes a reasonable request for assistance at the forefront of the business of the call. Rather than the caller unproblematically requesting help, which the call taker can then grant or decline, the “not an emergency” preface raises the question of whether the call should have been made and holds it up for the call taker’s evaluation and judgment.

Our analysis that “not an emergency” prefaced reasons for calling are designed to mitigate the caller’s responsibility for claiming an emergency can be exemplified by the cases in our collection that explicitly refer to third parties prompting the caller to make the emergency call. In Extract 8, the caller orients to an incipient emergency and states that she has permission to call “before violence started” (line 9).

Extract 8. 911 Domestic

The caller orients to an incipient or potential developing emergency by inserting “immediate” into the “not an emergency” preface (line 2). In so doing, she orients to her decision to call as potentially premature and unwarranted. Like most other cases of this phenomenon, the caller then provides a description highlighting emergency implicative elements of an active and nontrivial (tremendous) domestic dispute. It justifies the caller’s grounds for concern because her daughter is involved (line 5). With the call taker’s assent to continue the account (continuer on line 7), the caller justifies the potentially premature call by stating that she is following the instructions of a police officer who, presumably after a previous incident, encouraged such preemptive contact. The call taker treats this construction as a prima facie acceptable reason for calling as they then engage in incident-relevant questions (line 11) rather than challenging the call’s legitimacy.

A striking feature of our collection is that the caller’s first formulations almost exclusively contain reporting or descriptive formats (e.g., “I wanna report,” Extract 2, line 5). Kent and Antaki (2020) have previously described distributions of requesting, reporting, and descriptive formats for first formulations in emergency calls. Our collection’s extreme paucity of requests is atypical. Describing is a different action from making a request. Whereas requests encode a display of the speaker’s entitlement to expect the request to be granted (Curl & Drew, 2008; Heinemann, 2006), descriptions provide information about a situation of putative relevance to the recipient and leave unstated the evaluation of what response is appropriate.

Extract 9 is the only example in our core collection that contains an explicit request at the end of a situation description. Unlike our other examples, the description does not highlight features consistent with a candidate emergency or leave the assessment up to the call taker. Instead, the caller’s grounds for expecting assistance are asserted through an explicit request, “I’d like a police officer please” (lines 6–7). This causes interactional difficulty when the request is refused for not meeting the categorical threshold for police action.

Extract 9. 999 Barred from McDonald’s

Upon phoning 999 (UK emergency), the caller immediately acknowledges “it’s not >really e< mergency” (line 3) and then delivers the description of the situation that prompted the call (lines 4–6). Unlike other cases, it is hard to find any aspects of the description that index potential emergency elements: The caller has been denied access to a restaurant due to intoxication. Instead of maximizing the emergency implicative elements of the description, the caller tags on a request, “I’d like a police officer please” (lines 6–7). In so doing, they anchor their reason for seeking police assistance in their entitlement to expect such requests will be granted rather than in describing a situation that objectively meets the categorical threshold for an emergency. Additionally, the caller’s delivery indexes potential intoxication, which could undermine their status as a “competent member” of the interaction and cast further doubt on their capacity to accurately categorize an emergency incident.²

When the request for assistance is refused, the call taker justifies their refusal because the situation falls outside the police’s remit (“It’s not a matter for the police,” line 19). This asserts their judgment that the call did not meet the categorical threshold of a police emergency (or even nonemergency police involvement). The formulation of the decision does not address the caller’s claimed entitlement to request police assistance, which might account for why the caller subsequently upgrades the request’s entitlement with an “I want … ” formulation (line 20) and roots it in their innocent victimhood by having “done nothing wrong” (line 21) and being unjustly prevented from going about their evening.

Not taken at face value

Our final analytic observation is that there are no examples in our collection in which the call taker takes “not an emergency” to mean the call is not about an emergency and terminates the call or rebukes the caller for misusing the emergency number. Other cases from our datasets (e.g., Extract 1) reveal that call takers are not reluctant to challenge callers quickly and decisively for failing to promptly demonstrate the attributes of their situation that justify the emergency categorization. For “not an emergency” prefaced calls, preemptively acknowledging the borderline emergency categorization of the call might buffer callers against a swift reprimand for a call that ultimately does not require a police response and affords the caller some leeway to stretch the normative sequential expectation that they present a candidate emergency as their reason for calling. In Extract 10, the caller begins with “not an emergency” but does not go on to counterbalance that statement by describing a situation containing markers consistent with a candidate emergency.

Extract 10. 911 Getting down with the world

The caller begins his first turn with “It’s not an emergency yet” (line 4). However, what follows is not a description of the situation. “It can become one” (line 5) instead orients to the foreseeable potential of their, as yet unstated, circumstances to deteriorate into an emergency. The call taker’s repair initiator (“You have. a wha:t?”; line 6) makes relevant a response from the caller to classify his situation with a description of what he is dealing with that would match the category of an emergency. It reveals an orientation by the call taker that the caller should justify his call by outlining a candidate emergency.

The caller resists the push for a concise formulation of his reason for calling and requests time to “build up to” (line 10) his description. He repeats the “not an emergency” + “could become one” formulation (lines 10–11) again without any markers (urgency, seriousness, risk of harm) shown in earlier extracts. All the caller has reported is the potential for an unspecified situation to become an emergency within an unspecified timeframe. The caller delivers an extended account (omitted lines 14–28) of nonacute difficulties experienced by a friend culminating in a current state of being “down with the world” (lines 29–30). On line 31, the call taker repeats his earlier (line 12) request for the caller’s location. This sequentially deletes the extended narrative (Schegloff, 2007), marks it as inconsequential for the progression of the emergency call and reasserts the original question. Of relevance here is the retention of the location focus to the call taker’s question (and its implicit association with preparing for emergency dispatch) instead of querying the legitimacy of the call through a more skeptical question design (Kent & Antaki, 2020). Once again, the caller requests additional time by imploring the call taker to “listen” and to “wait” while he finishes his narrative before answering the question (lines 32–33). This displays an orientation by the caller to the nonnormative nature of taking an extended time to provide the reason for calling (Whalen et al., 1988).

The call taker highlights the importance of the caller quickly providing information consistent with an emergency to justify continuing the conversation instead of responding to the three waiting emergency calls (line 35). Thus, the call taker reprimands the caller for not efficiently justifying his reason for calling emergency assistance. The caller dramatically rejects the call taker’s demands (line 37), announces, “I’m jus’ gunna kill myself” (line 39), and immediately disconnects the call. This is the first time the caller has provided information that could constitute an emergency. In this position, it is not done to report circumstances for the call taker’s evaluation as a candidate emergency (as is the case with the other calls in our collection). The caller displays expletive-laden frustration that his attempt to control the progressivity of the conversation has been undermined by the call taker’s adherence to the normative structural organization and constraints within an emergency call despite his preemptive acknowledgment that their situation was “not an emergency.”

Stating that something is “not an emergency” does not remove the structural expectation that an emergency-relevant description should follow. It is not a carte blanche for callers to remove the constraints of calling an emergency line. It simply marks that a caller positions their situation as sitting on the boundary of what might constitute an emergency and offers the situation to the call taker for their evaluation. Doing this type of report with the “not an emergency” preface helps exonerate callers from the responsibility of making the independent decision that emergency service will help resolve the problem. In doing so, callers leave it up to the call taker to assess the situation as requiring emergency assistance.

Discussion

“Not an emergency” constructions are a rare but robust phenomenon when calling emergency services. They are predominantly located immediately following the call taker’s answer to the emergency call summons and are the first part of a counter-balanced turn design in which the second part is a descriptive account of the reason for calling that indexes features of the situation that could constitute a (borderline) emergency. Reasons for calling formulated in this way appear designed to position the call as a boundary case for qualifying as a legitimate emergency, often due to its imminent potential to cause damage or injury. They prospectively manage the caller’s responsibility for potentially making a frivolous emergency call by displaying their own orientation to this risk and, rather than requesting assistance, report a candidate emergency for the call taker’s evaluation.

A key orientation for callers using “not an emergency” prefaces was the management of their deontic entitlement to expect police assistance. In our data, “not an emergency” almost exclusively prefaced descriptive formulations in which a caller chooses to inform the call taker of their situation rather than explicitly request assistance. This work contributes to existing literature on entitlement and deontic rights to control and make decisions about future courses of action in interaction (Stevanovic & Peräkylä, 2012). It highlights the utility of descriptive formulations for avoiding asserting of one’s own deontic rights and, instead, placing that responsibility on one’s interlocutor. Here our work aligns with observations by González-Martínez and Drew (2021), who, when studying nurses’ intrahospital telephone, calls found that informings can be a way of recruiting assistance on the basis of the recipient’s responsibilities toward the subject of the informing; in their case, patients’ readiness to transfer to a different ward within a hospital. For calls to emergency services about situations with ambiguous emergency characteristics, descriptive formulations (especially when prefaced with “not an emergency”) privilege and foreground the call taker’s rights and responsibilities to ensure that candidate emergencies are not prematurely dismissed as ineligible.

We showed how the interactional construction of what counts as an emergency is connected to the interactional categories of caller and call taker and their respective social responsibilities as “citizen” and “police.” It contributes to a growing picture of the role of categorical practices in the organization of social encounters (Watson, 2015). Turowetz and Maynard (2016) demonstrated the ongoing process of categorical attribution across an extended autism diagnosis consultation. Here, we have explored a much more acute context in which the soundness of a presumptive categorical attribution (emergency) is queried, evaluated, and (in the majority of our cases) ultimately accepted. In particular, our analysis demonstrates the methodological potential of approaching the study of categorization in interaction from a starting point other than personal identity memberships. In so doing we hope it adds to the comparatively impoverished body of work on activity categorization (e.g., how are the limits of what constitutes an emergency constructed and negotiated in interaction).

Emergency (or not) categorization has conceptual parallels with diagnostic labels. Both categories, when accepted by service seekers and providers (emergency or medical), grant access to resources and expertise that would otherwise be withheld. Horton-Salway (2007) examined how doctors and patients attempt to construct a legitimate category for myalgic encephalomyelitis (ME/chronic fatigue yndrome) and “police” its boundaries, often in contrast to related (psychological) symptoms that might not achieve the ME categorical diagnostic threshold. Although our analysis is organized around a designedly counterbalanced (or contrastive) turn design, it is striking that similar practices for constructing contrast have been noted in other institutional settings where high stakes categorical determinations are in play—for example, courtroom cross-examinations (Drew, 1992) and medical diagnoses (Horton-Salway, 2007; Turowetz & Maynard, 2016).

Our analysis highlights one, fairly simple interactional moment in which the complex nuances of an individual’s circumstances are juxtaposed with rigid (but only implicitly defined) boundaries and thresholds for accessing assistance from emergency services. At risk in such cases is the caller’s moral status as a “good citizen” and the (in)adequacy of the emergency category as a threshold for action if it cannot be accurately judged by those who call for help. How callers orient to and manage the possibility that their call will not result in the assistance they seek is obviously a far more pervasive concern than just for the rare callers who choose to use a “not an emergency” preface (e.g., Whalen & Zimmerman, 1990). However, the stark incongruity created by the phrase presents a vivid illustration of the intersection between categorization and action in institutional interaction. It is striking that attempts to gain assistance for situations that are only marginally institutionally relevant might actually be strengthened by foregrounding rather than disguising the imperfect categorical attribution. We feel this warrants further research attention.

Disclosure statement

The authors confirm that there are no relevant financial or nonfinancial competing interests to report. The datasets for this data are not publicly available due to their sensitivity. However, both authors welcome contact from other researchers who might like to explore the data further.

Notes

¹ National roadside recovery organisation.

² Clubbing venue.