Auras of Detection: Power and Knowledge in Drug Prohibition

Methods and Mess

Emily uses a small metal spatula to scrape around the inside of a used cooker, taps the flakey residue out onto a tiny square of foil, and positions it in the center of the machine. When she starts the scan, a bevy of colorful lines populate her screen. These, she tells me, are spectra. This portable mass spectrometer measures the vibrations of the molecular bonds. Those wavelengths are then rendered graphically on the screen, with tell-tale peaks indicating the presence of…something, in the sample.

Emily starts pulling up electronic libraries of known substances, overlaying spectra of known molecular compounds on the sample’s output, hoping to find a reasonable match “It’s just what you think fits best,” she says. The fentanyl analog they’ve been seeing in samples, which she expects to find here, is identifiable by a single, predictable peak. Since the actual quantity of fentanyl in illicit opioids is relatively small, any amount that’s in the sample may be too small for the machine to detect. In that case, the output on Emily’s screen would simply show the signal derived from other, bulkier ingredients. “We think about fentanyl being the main ingredient in the dope,” she says. “It’s actually such a small part of the product, though, so it can be hard to find.”

Emily possesses an encyclopedic knowledge of the local drug supply, but she is still learning to use this machine. She spends a significant amount of her time arguing with Matt, a chemist who has been providing technical assistance, an expert in the interpretation of spectrometry data, and someone who knows practically nothing about drugs. Throughout the day, they amicably quibble back and forth about the reasonableness of certain interpretations, with Matt’s significant technical expertise often running afoul of Emily’s street smarts, and vice versa. On this day, a debate ensued about whether a particular curve shown by the spectrometer could have indicated licorice root, as the software suggested, in a local drug sample.

“It matches here, here, and here,” Matt observed, pointing to the screen.

“Matt. It’s not f---ing licorice,” Emily sighed.

Later, when the machine was being packed up for the day, Emily mused, “I’ve compared running local dope through the machine to analyzing cake. In general, you know what you expect to find. There’s flour, there’s some eggs, sugar. But the information you really want is the leavening agent. Was it yeast? Was it baking powder? Was it nothing at all? It’s the most powerful ingredient in the cake, but it’s there in such tiny amounts that it gets drowned out by all the other ingredients. Figuring it out is more of an art, than a skill.”

The data and analysis presented in this article come from a patchwork of experiences that span multiple years, regions, research projects, and professional partnerships. Systematically collected data comes from two formal research studies. ¹ First, in 2016 and 2017, I interviewed more than 100 residents of Providence, Rhode Island as part of an exploratory-sequential study (Fetters et al., 2013) that combined ethnographic interviewing and flexible survey methods with grounded analysis (Glaser & Strauss, 1967) to investigate how people who use drugs were perceiving and responding to the influx of fentanyl into the local drug supply. In brief, interview transcripts were inductively analyzed, hypotheses generated, and interview guides amended to pursue emerging hypotheses on an ongoing basis for a period of 15 months.

Later, starting in 2018, two colleagues and I began a multi-site study that used the Consolidated Framework for Implementation Research (Damschroder et al., 2009) to evaluate the provision of novel harm reduction strategies for overdose prevention. Over a 2-year period, we made numerous site visits, conducted hours of in vivo observations, and interviewed dozens of staff persons and clients at participating harm reduction organizations. Most of these programs offered some kind of drug checking service, either as the core of or as an important adjunct to their client engagement strategies. These numerous notes and transcripts were analyzed inductively through thematic coding (focusing especially on emergent themes related to the role of drug checking technologies in knowledge production about tested drug products) and deductively using the 36 constructs included in the Consolidated Framework for Implementation Research, which aims to isolate known barriers and facilitators of intervention implementation in novel settings (Damschroder et al., 2009).

Beyond these formal research projects, I have spent the past 4 years consulting for the public health arm of the Overdose Response Strategy (ORS), a federal public health/law enforcement partnership between the CDC (the United States’ national public health agency) and High Intensity Drug Trafficking Areas (HIDTAs; a drug enforcement program run by the U.S. Office of National Drug Control Policy). In this role, I have gradually become oriented to the world of policing and the biopolitical logics that drive coordinated efforts of drug law enforcement at the federal, state, and local levels. Overtime, through numerous planning meetings, emails, phone calls, and visits to the hotel bar between conference sessions with various ORS collaborators, I became familiar with some of the many strategies law enforcement professionals can engage to produce knowledge of and continuously validate their epistemic authority over the illicit drug market. Hardly systematic, my insights about the cultures, strategies, and discourses of law enforcement institutions emerged organically though the repetitive motions of my professional duties as a public health consultant: explaining public health to law enforcement; explaining law enforcement to public health; and explaining my research about law enforcement (Carroll, Mital, et al., 2020) back to law enforcement to facilitate meaningful public health partnerships. I refined my understanding of my interlocutors’ conceptual schemas and modeled worlds every step of the way.

In reference to the insights of STS scholars Bruno Latour and Steve Woolgar, John Law has described the production of scholarly knowledge as a phenomenon that takes place “in a more or less messy set of practical contingencies” (Law, 2004, p. 13). Similarly, I situate this analysis in the mess from which it emerged—in the uneven and unkempt ways in which the social imaginaries produced and perpetuated by the War on Drugs take shape in the world. The textures of my knowledge about the two domains that occupy the focus of this paper are, therefore, uneven: one (of people who use drugs) based on systematic research, partnership, and advocacy; the other (of law enforcement) emerging from a place located somewhere between ethnography and pedagogy. Both are grounded methods in their own way, but they are shaped by different aims, contexts, power structures, and ethical imperatives.

In that sense, much like Emily’s description of drawing conclusions from the mess of information displayed by her mass spectrometry software, my fluency in the discourses and sociotechnical complexities of knowledge regimes about the illicit drug market has been gained through a variety of pathways, from the highly structured to the familiar and informal. Theoretical clarity emerged over time in a way not unlike Sheila Jasanoff’s description of analyzing messy data: “…making sense of complexity [and] discovering stories that matter in a field of infinite happenings, as when a world of randomly colored points resolves itself into a figure and ground through the artistry of the pointillist painter” (Jasanoff, 2017, p. 2). All the while, my own ontological politics (namely, my firm commitment to the idea that people who use drugs possess valuable expertise that deserves to be acknowledge and accounted for) acts as a guard rail, providing structure to my analysis without forcing a tidy point of view that has been “distorted into clarity” (Law, 2004, p. 2).

In this analysis, then, I bring these multiple circumstances into conversation with each other to provide a view on how political power, knowledge regimes, and access to technological instruments of detection intersect to shape what we know—and what we think we know about what we know—about the illicit drug market.

The Things We Can See

Vanessa was recruited by a harm reduction organization to provide outreach services in Northampton after a few years receiving those services herself. She came to this area from New York, where the practice of using stamps to brand and distinguish illicit drug products from different suppliers was still common, as it was in this region when she first arrived.

“A while back,” she recalled, “there were stamped bags of [illicit] methadone that were dropping people left and right. Then somebody in Holyoke [started using] the stamp and somebody in Springfield [started using] the stamp, and the Holyoke bags were high [body] count. And I kept thinking how—This isn’t Walgreen’s, you know? Like, I am so privileged to walk into Walgreen’s and get a little bottle that reads ‘Vanessa,’ this many milligrams of this sh—, and that’s what’s in there. Why can’t people who use drugs, like, why are we not afforded that right? Like I have the right to know what I’m putting in my body, anything that goes in my body, and because I choose to use drugs, I don’t get that right? Or even a fentanyl test strip?…Why don’t we get drug [checking] [i.e. for safety] Why are the police doing drug [checking] [i.e. for criminal investigations]?”

Listening to her, I recalled the strategies for discerning the contents of illicit drugs that were shared with me by residents of Providence, Rhode Island. A variety of descriptive attributes (what color is it, what does it smell like, what does it taste like, how does the color or consistency change when it is cooked or pulled into the syringe) were used to discern the likely presence or absence of adulterants like fentanyl—with decidedly mixed success.

Vanessa lamented that the aggressive policing of drug distribution had effectively eliminated stamps from the local drug market in her region. She recalled an occasion when “a girl came in and brought me two bags. I was like, ‘Crap.’ I was like, ‘They don’t have any stamps on them.’ It was, like, just a blue bag and a yellow bag. And she was looking at me like my dog. I mean, like, she tilts her head like, ‘Why are you so stupid?’ And she’s like, ‘[It’s] because they’re tracing the stamps back. And, like, if you have [this stamp] and somebody OD’s, you’re getting charged.’ And now they’re not even stamping the bags, and so the branding—even the pitiful little branding information we had is now gone.”

The fundamental theoretical premise of STS studies is that objective information about the world is anything but. Bruno Latour illustrates this phenomenon well in his ethnographic study of soil scientists working in the Amazon rainforest (Latour, 1999). Watching them dissect a smooth spectrum of colors in the soil into a taxonomy of discrete hues corresponding to the color chips they had in hand, he concluded that scientific knowledge comes not from the collection of objective facts obtained through unmediated, face-to-face interaction with the world, but through a system of continuous translations between concrete material and abstract signs (Latour, 1999). We can think of the stamps by which drugs are branded or the taxonomies of colors and textures described above as abstractions in the same way—efforts to produce some kind of clarity out of mess.

Our epistemic frameworks—the systems of meaning and abstractions with which we produce clarity—tend to naturalize certain modes of thinking. They enable us to black box certain “facts” away from their social origins and perceive them a priori, as just “naturally” or “obviously” true. Epistemic frameworks can also naturalize certain ontological standpoints—social positions from which one can make claims about the world—as credible or valid. In her analysis of knowledge and power in policy making, Jasanoff identifies three such standpoints (which she refers to as “authorized…modes of [collective] seeing” or “regimes of sight”) in contemporary American society (Jasanoff, 2017). The first is the standpoint of scientific method, as it has been traditionally conceived. She calls this “the view from nowhere” in reference to the core (though not entirely true) political claim of scientific practice that it is, by its very nature, pure, objective, and beyond the influence of social and political contexts. Here, I think of Emily looking at the output from the portable mass spectrometer. The colored tangle of lines on her screen is understood as raw, unmediated, objective representation of the sample. If that information is not immediately clear, the problem is believed to exist in with the operator, not the machine.

The second regime Jasanoff calls “the view from everywhere,” which is authoritative in its ability to broadly apply reason and for its claim to reflect views that, if not universal, are “inclusive of all affected positions” (Jasanoff, 2017). Though Jasanoff conceives of this position as one held by, for example, expert advisory bodies who are deemed capable of synthesizing many aspects of a problem into a single, reliable whole, Vanessa also occupied this standpoint as she shared her expertise distinguishing illicit drug products and her experience-based reasoning about how people who use drugs are impacted as these practices change. Jasanoff’s third regime of sight is “the view from somewhere,” the subjective but authoritative reporting of the individual eyewitness. Dan and many others who described their lived experiences consuming Providence’s “weird heroin” adopt this standpoint, from which—no matter how unbelievable their claims might seem to be—we are compelled to take them seriously.

Central to this analysis is not only how certain ways of seeing become naturalized, but also how power operates through our epistemic frameworks and knowledge-production practices. Michel Foucault theorized knowledge and power as co-existing, mutually constituting one another in a kind of knowledge-power feedback loop. Socially powerful actors, he argued, possess the epistemic authority to produce and verify knowledge about the world, while, simultaneously, the production of knowledge in accordance with the intentions of the powerful serves to reify that position of authority (Foucault, 1990). In this same way, powerful social actors (medicine, public health, law enforcement) have, through varied means, established themselves as “authorized seers” of illicit drugs in their respective domains.

Further, epistemic authority does not always but is very useful for delimiting the standpoints from which people who use drugs are enabled to act as “authorized seers” and producers of knowledge (Jasanoff, 2017; Walker, 2021). In the United States, popular narratives about substance use and addiction cast people who use drugs as untrustworthy (Bourgois, 2000), unable to accurately know or report truthfully about themselves (Carr, 2010), and socially hazardous in their beliefs and behaviors (Garriott, 2011). Authoritative representations of people who use drugs offer a symbolic richness to this denial of credibility. Professional narratives about “drug seekers,” “dope fiends,” “poison pushers,” and “manipulators” solidify the social imaginary of people who use drugs as unreliable—or even dangerous—“others” who cannot accurately look upon the world or faithfully report back what they see. Indeed, this very logic has been used by federal mental health authorities in the United States to argue that providing people who use drugs with access to drug checking technologies is pointless as it is “based on the premise that a drug user is poised to use a drug is making rational choices…and is thinking completely logically about his or her drug use,” which, they insist, “could not be further from the truth” (McCance-Katz, 2018).

In the sections that follow, I discuss how law enforcement adopts certain “regimes of sight” to enact and validate their epistemic authority over the illicit drug market. I then turn to the social and epistemological impacts of drug checking technologies in the hands of people like Dan, Emily, and Vanessa, exploring this new intervention as an act of resistance.

Collective Seeing From Nowhere and From Everywhere

In the fall of 2018, I was invited to Washington D.C. to participate in a workshop for drug policy experts. The afternoon was dedicated to panel presentations. When representatives from law enforcement agencies were at the dais, someone in the audience posed a question about “fentanyl-laced marijuana.” Namely, had the panelists ever seen such a thing on the job, as rumors of its existence had been circulating but had never been verified.

One speaker, a representative from a mid-sized urban police department, answered in the affirmative. He proceeded to recount an incident when a fellow officer made an arrest and found marijuana among the arrestee’s possessions. He explained, “There were several buds of marijuana in a disposable plastic water bottle. The officer, sort of as a reflex—’cause who knows why we do what we do sometimes—unscrewed the lid and sniffed it. And wouldn’t you know, there was fentanyl. And, just, boom.” He snapped his fingers for dramatic effect. “He fell over and started to overdose.”

The speaker must have seen the disbelief plastered across my face, because his next move was to turn to me and say, “Really! I can’t make this stuff up!”

“Well. Someone did,” I replied. ²

My suggestion that these events could not have happened as described was met with objections from some (though not all) of the other law enforcement professionals in the room. Many claimed to have personally witnessed fellow officers experience an unexpected overdose just like this one. The discussion moderator took interest in our disagreement. He pressed those who reported witnessing their colleagues overdose. Was the officer taken to the hospital? Were any toxicology tests performed? Was fentanyl exposure ever confirmed? Did the officer lose consciousness, stop breathing, or display any other definitive symptoms characteristic of opioid overdose? Universally, they had not.

By the time the day had drawn to a close, every person in the room who had earlier claimed to have witnessed an officer overdose due following exposure to fentanyl in the field had willingly conceded that their colleagues had, in all likelihood, experienced a severe panic attack, not opioid toxicity.

In contemporary American society, law enforcement institutions maintain far-reaching epistemic authority over the illicit drug market. From organized raids of “Chinese” opium dens in the 1800s to the over-incarceration Black Americans in the “crack epidemic” of the 1980s to today's “kinder, gentler drug war” against a largely white population of people who use drugs, the policing of drug markets in the United States has always been a vehicle through which state powers pursued broader, racialized political agendas (Alexander, 2013; Beckett & Brydolf-Horwitz, 2020; Campbell, 2007). Today, in the thick of political “colorblindness” that obscures the centrality of race in our imaginations of drug use and drug criminalization (Alexander, 2013; Hansen et al., 2020), the enforcement of drug laws also tends to appear stripped of its racialized logics. Instead, drug law enforcement has increasingly become framed as a complicated technical endeavor, one that William Garriott has called “narcopolitics,” (Garriott, 2011), focused not on policing people, per se, but on controlling the circulation of illicit drugs by, in turn, policing those who circulate them. This view was laid out quite simply by an officer with the U.S. Drug Enforcement Administration (DEA) whom I consulted in 2017 about his role as a federal overseer of medical clinics that provide buprenorphine, an effective and lifesaving medication for opioid use disorder. “I don’t care how [the doctors] choose to treat their patients,” he said. “All I care about is that the [buprenorphine] makes it into the bodies it was intended for and doesn’t end up somewhere else.” In other words, he was not interested in how—or even how much—clinicians chose to prescribe. His only concern was the drugs that circulated outside of bodies and thus constituted a piece of the illicit drug market.

In my systematic research and in my professional experience, I have observed two ontological standpoints frequently adopted by representatives of law enforcement institutions in order to validate this epistemic authority over illicit (or illicitly circulating) drugs: a scientific “view from nowhere” and an expert “view from everywhere.” The first of these, the scientific “view from nowhere,” operates by appealing to the aura of objectivity surrounding advanced drug detection technologies. Take, for example, the 2019 testimony of Chief Postal Inspector Gary R. Barksdale before the U.S. House of Representatives. He said, “Winning the battle against illicit drugs in the mail stream is a top priority for the Postal Service and the Inspection Service.” As Chief Postal Inspector, he said, he aims “not only to remove drugs from the mail but to investigate and arrest individuals and disrupt drug trafficking networks across the U.S.” Barksdale further reported that the Inspection Service had “conducted investigations using cyber and analytic capabilities and invested in advanced technology.” He reported equipping field inspectors with TruNarc® devices, hand-held Raman spectroscopy narcotics analyzers, presenting this development as an institutional achievement “enabling our inspectors to scan more than 450 suspected controlled substances in a single, definitive test” (Barksdale, 2019). By speaking through this technological register, Chief Inspector Barksdale effectively folded the language of science (“scan,” “test,” “investigate”), technology (“advanced,” “analytic,” “cyber”), and warfare (“battle,” “seize,” “arrest”) into a narrative framework that posits illicit drug markets as a fundamentally technical problem that law enforcement actors are uniquely skilled and technologically equipped to resolve. That technology, frequently used by law enforcement yet legally classified as drug paraphernalia across much of the United States when in the hands of people who use drugs, validates both the epistemic authority of law enforcement over the drug market as well as the policing powers they possess to disrupt it.

In an article dissecting how the second common ontological standpoint, the “view-from-everywhere,” is often established and validated in courtrooms, Sheila Jasanoff (2018) provides a helpful metaphor that compares the expertise of an aeronautics engineer to that of a beekeeper. In this metaphor, the engineer is someone who generates knowledge by applying scientific reason. They can explain how bees are able to fly using theoretical principles. Alternatively, the beekeeper can provide expertise based on countless iterations of lived experience with the bees. They can offer insights like, “bumblebees always take off into the wind.” Both the engineer and the beekeeper are experts. But, as noted in this and other court decisions referenced by Jasanoff, “[while] the beekeeper does not know any more about flight principles than the jurors…he has seen a lot more bumblebees than they have” (Jasanoff, 2018, p. 18).

It is not uncommon to see law enforcement framed as an institution composed of experts like the engineer. Consider, for example, the consistent over-representation of law enforcement leadership as plenary speakers at the annual Rx Drug Abuse and Heroin Summit, which bills itself as “the largest annual gathering for stakeholders to discuss what’s working in [drug use] prevention and treatment” (Rx Summit, 2021). Why would law enforcement be invited—or even eager—to speak about drug markets to interdisciplinary audiences unless their expertise operated at the level of broadly applicable theoretical principles? At the same time, in a more quotidian fashion, law enforcement professionals may embody the expertise of the beekeeper as circumstances demand. Consider, for example, the response I evoked from a police officer in 2017 when I asked how his department was responding to rising overdose rates: “I don’t view drug use as a sickness,” he said. “It’s not a disease, it’s a choice…and that’s not from being stupid, it’s from 25 years of policing.” The authority of this officer’s claim rests on the unspoken presumption that, compared to the rest of us, he has seen an awful lot more bumblebees.

Over time, I have come to understand the social importance of this “view from everywhere” in the style of the beekeeper as a method for law enforcement actors speak with finality—to produce knowledge about the world that reifies their authority within it. The countless reports in mainstream media about officers overdosing after encountering fentanyl in the field, like the story I and other workshop attendants were told in D.C., is a robust example of this phenomenon, typically relying on the acceptance and reproduction of unlikely claims made by law enforcement representatives as objective truth. The best-known origin of this risk narrative dates back to 2016, when the DEA released a roll call training video that told police officers, “You could be in grave danger if you come into contact with fentanyl” (U.S. Drug Enforcement Administration, 2017). The explicit message that fentanyl could be deadly for officers was articulated along-side images of field teams in hazmat suits and video clips of two officers testifying to their own alleged overdose experiences after incidental fentanyl exposure in the field. These two officers described symptoms incongruent with opioid toxicity, but the message still hit its mark. Since then, reports that officers have overdosed and even required the overdose reversal medication naloxone to survive have been steadily reported in local and national media (Anderson, 2017; Pollauf & Kuehnlein, 2018; WCVB-TV, 2019). Even within the past year, local news stations in the United States have aired interview segments with police representatives who state, unquestioned, that officers handling unknown substances have been “put [into] an overdose where they had to receive Narcan® [naloxone] spray” (Winn, 2020) and that “if you even touch [synthetic opioids] you could die” (Lofaso, 2020).

According to a statement from the American College of Medical Toxicology and the American Academy of Clinical Toxicology, overdose resulting from incidental skin contact with fentanyl or any other synthetic opioid is practically impossible (Moss et al., 2017). Further, a systematic review conducted in 2020 found that no single report of an officer overdosing in the field could be reasonably confirmed (Herman et al., 2020). In keeping with my experience in D.C., most case reports available list symptoms congruent with severe panic attacks, not opioid toxicity (Attaway et al., 2021). In response to this counterfactual information, government agencies in Canada and the United Kingdom have retracted guidance indicating that incidental exposure to fentanyl poses a meaningful safety risk (Public Health England, 2018; Tunney, 2019). Some American law enforcement agencies, by contrast, have continued to make implausible claims in support of the idea that fentanyl presents a serious danger to first responders, including allegations that officers experiencing potentially fatal overdose had survived the ordeal by administering naloxone to themselves (Santa Rosa Police Department, 2020).

Years ago, I was working full-time with the ORS, there was no better way to clear a room than to suggest that these reported overdose events were not actually overdoses, or that the risk allegedly faced by first responders from illicit drugs on scene might be minimal. Even my suggestions that it is fundamentally cruel to officers to perpetuate the false belief that they are in mortal danger were not well received. I do not doubt that the officers who declared fentanyl a deadly risk for first responders were expressing their genuine beliefs. After all, the source of their information—the DEA—was a law enforcement institution with supreme epistemic authority over illicit drugs, even compared to other law enforcement agencies. But the tenacity of those beliefs and the temerity with which I have seen officers defend those beliefs in the face of strong counterfactual information, I contend, is significant. It suggests that more is at stake in these debates than the sheer “facts” about fentanyl.

The durable commitment I have seen some law enforcement professionals display to the conclusion that these events are, in fact, potentially fatal overdoses reveals these claims to be something of a ritual in collective seeing, in collectively shaping the world and presenting it to others with epistemic authority. In D.C., the chorus of “It’s happened in my department,” “And mine,” “It’s happened to my officers, absolutely,” in response to our conference moderator’s questions rang out like a well-known hymn. In news reports that detail alleged overdose events, this ritual of collective seeing is reinforced, the epistemic authority of law enforcement ensconced in the public record as journalists uncritically echo local officers’ reports and fail to ask whether these claims are verifiable or even reasonable (Siegel, 2017). Thus, questioning this authoritatively posited set of “facts” not only threatens the soundness of those “facts,” it also threatens the epistemic authority of law enforcement through which those “facts” are presented as truth.

Considering these two ontological strategies—adopting the technologically authoritative “view from nowhere” and the socially authoritative “view from everywhere”—we can imagine how drug checking technologies might occupy a place of moral ambiguity when they are posited as tools for overdose prevention. As noted above, technologies for drug analysis—which validate law enforcement’s epistemic authority when securely placed in law enforcement’s hands—could be construed as undermining efforts to police the drug market through interdiction and disruption when placed in the hands of people who use drugs. They are classified in some state laws and perceived by many law enforcement professionals as drug paraphernalia. Tolerating the use of these tools for harm reduction purposes, as one law enforcement partner recently described, requires one “to be resigned to the fact that the illegal drug use is going to happen.” For some, that will be a shade too far. Consider, also, how the policing of drug markets in Northampton, where Vanessa works, deterred the use of stamps and, thereby, markedly reduced the capacity of local people who use drugs to discern one illicit drug product from another. For law enforcement, their own capacity to disrupt drug sales clearly outvalued the capacity of people who use drugs to develop meaningful knowledge about the local drug supply.

How, then, does the meaning of drug checking technologies, especially advanced technologies for drug analysis, change when they are removed from the sole purview of law enforcement and placed in the hands of those whose ontological standpoint is not valued? How will law enforcement institutions and actors within those institutions respond to the re-orientation of those technologies away from interdiction and toward preserving the safety of people who use drugs? I turn, below, to the experience of a few harm reduction organizations at the vanguard of advanced drug checking services to explore answers to these questions that are emerging from their work.

The FDA of Siler Street

Nathan and Brittany, who were on shift at this harm reduction organization, just stopped for lunch. Their new COVID-19 operating procedures dictate that all supplies (like syringes and naloxone) be distributed outside, leaving the drop-in center eerily quiet at this time of day. I was there as an observer, deeply curious about their new drug checking service. I listened intently to their reflections as we all greedily tucked into our meals.

“But, like, honestly, how do you know you have control?” said Nathan, punctuating his question with a flourish of his sandwich. “How do you know that you’re—that you could be in control as a consumer? If you just need to get whatever’s closest to you whenever you can get it? But if you have the ability to check and know that, oh, the guy that shows up in the plaza at 3 o’clock has stuff that’s not going to ruin my life. I might still make my appointments tomorrow if I get that stuff.”

Brittany mused, “Yea, so, interestingly, we had a couple of—actually one recently that I totally forgot about is, where this guy was like, ‘I had, like, an out of body, like, seizure-like experience.’ And I didn’t really see anything [on the spectrometer], so we sent it in [to a bigger lab for confirmation] and it was a synthetic cannabinoid…The police down south were saying they were finding weird cannabinoids. And I was like, ‘Nah.’ First, I don’t believe anything the cops tell us. But then, like, come to find out they actually had some data to back it up. And we found the synthetic cannabinoid in some samples. And I was like ‘What the f---.’ Basically, the more we do this the more I realize I don’t know sh--. But, like Nathan said, one of my favorite things to share with people is that we will sometimes work with people who are selling, and they’re so motivated and so excited to be able to, like, know what they’re providing to consumers and making sure that it’s safe, and that is something I find to be so rewarding.”

“That would be key, right?” said Nathan. “Getting a dealer to know exactly what’s in there so they could build trust with a buyer…Most people do seem to trust their dealer. Like, he’s a regular person. But then my next step in that conversation is, yea, but who’s their guy? And who’s that guy’s guy? And who’s that guy’s guy? And that guy’s guy? Like, have you met them? Cause that bag stopped eight times since it was a brick, and it’s cut in every spot. But this intervention of giving people information about what they’re putting in their body—like we have an entire f---ing government agency dedicated to that, the FDA. Do you get what I’m saying? So, like, why would we not value that for drug users who are taking very serious risks?”

“Right? We’re the motherf---ing FDA of Siler Street!”

Criminalized markets are de facto unregulated markets. Regulation creates opportunities to standardize formulations, ensure product safety, and provide consumers with meaningful information about ingredients and dosage. Without that information, illicit drug products are inherently uncertain commodities (as Vanessa noted, “This isn’t Walgreen’s, right?”). Further, the policing of local drug markets increases each individual participant’s risk of arrest, prosecution, or conviction. This added pressure leads many market actors to shift risk away from themselves and toward others, such as engaging in violence or threatening others in an attempt to restrict their behaviors and avoid drawing law enforcement attention (Dickinson, 2017). Policing also renders communication between actors across the illicit drug market inherently dangerous. The shift away from using stamps to distinguish illicit drug products, also described by Vanessa, is a direct response to that risk. This eliminates the opportunity for the development of shared categorical schemas about drug products across market actors (Cambrosio et al., 2006).

Over the past few years, I have seen people who use drugs employ a variety of strategies to mitigate this uncertainty. The most common of these, in my observation, is the development of taxonomies to compare illicit drug products to each other according to their descriptive characteristics (“dope that looks/smells this way or that way”) and phenomenological effects (“dope that feels this way or that way”). These taxonomies are largely individual and can vary significantly from person to person—thus explaining some of the danger inherent in purchasing drugs from someone unfamiliar (Carroll, Rich, & Green, 2020). Buyer and seller may have two very different ideas of what constitutes a “pretty strong” or “not that strong” opioid product simply because they have no safe opportunity to talk to each other about how their experiences line up. Some people do talk to others about their experiences with illicit drug products. However, criminalization actively deters people from sharing too much information. Most whom I have interviewed are able to have such conversations with a close relation such as best friend or a romantic partner. Occasionally, people can have conversations about the composition and effects of illicit drug products with someone from whom they regularly buy, but such exchanges require an extraordinary amount of trust, and trust can be hard to come by.

Drug checking services help people who use drugs chip away at these uncertainties. The first—and now most widely used—drug checking technology is fentanyl test strips: disposable, immunoassay strips that provide an accurate test for the presence of fentanyl with a low limit of detection (meaning it can, when used correctly, return a reliable result even for very small concentrations of fentanyl) that cost about one dollar apiece (Green et al., 2020). These tools allow people who use drugs to determine whether fentanyl is present in their drugs but do not provide information about the relative concentration of fentanyl or what other ingredients may also be present. More recently, a small number of harm reduction organizations in the United States and Canada with the means to spend $50,000–$80,000 on a device began piloting the use of portable high-pressure mass spectrometry (HPMS) and Fourier-transform infrared (FTIR) spectroscopy machines. These devices use slightly different mechanisms to return incredibly detailed results about the composition of drug products, including active ingredients like heroin, hydrocodone, fentanyl, and cocaine, as well as inert ingredients used as cut, such as mannitol, acetaminophen, or levamisole.

Many harm reduction advocates hoped that these technologies would perfectly inoculate their clients against uncertainty in that market, that they would be sufficiently advanced to return clear, complete, and unassailably accurate information about illicit drug products and produce what Jasanoff describes as “a view from nowhere,” collective seeing through purely objective technical registers (Jasanoff, 2017). If achieved, this would have fully eliminated the need for people who use drugs to develop messy and unreliable folk taxonomies based on color, smell, and feel. But the devices fell short of this perfect vision. Brittany explained it to me this way:

You need a visible sample for the FTIR, and its detection limit is like 5%. So, it doesn’t do trace amounts, right? So, in order to do trace amounts, which sometimes all we have is a trace amount [of a sample], the [HPMS] is helpful. But also, like, you can’t look at the FTIR or the spectra and be like “Oh I can clearly tell that this is what it is,” like, it takes a lot of expertise so, sometimes to cross reference both machines is the only way to do it. I’m just learning that there is no machine that’s perfect for this.

Yet, even though available drug checking instruments did not live up to these imaginations, the use of these machines does appear to facilitate the circulation of information about illicit drug products—and therefore the coordination of categorical schemas—across a wider field of actors within the illicit market. Put more simply, even though drug checking could not enable a uniform “view from nowhere,” it does enable a different kind of collective seeing—one that transforms isolated bits of knowledge and experience held by countless individuals into an ordered framework that is portable, transferrable, and widely-shared. In the words of Latour, drug checking enables new chains of abstraction to develop and circulate (Latour, 1999). The technology becomes the cross-network communicator that people who use drugs can’t safely be for themselves.

One way I have seen this coordination happen is by harm reduction staff incorporating what knowledge people who use drugs are able to produce for themselves into their interpretation of the HPMS and FTIR outputs. Emily, for example, is fond of saying, “You can’t ignore the input that people give you.” She might know enough on her own to be sure that a sample from the local drug supply doesn’t contain licorice, but producing a clear determination based on FTIR or HPMS output alone can be a gamble. Additional data about the phenomenological effects of the drug can add much needed guidance to her interpretation. She observed:

We definitely have gotten things where people [say], “I shot dope and I felt super racy.” And then [the HPMS shows] there’s a bunch of caffeine in it. And we’re like, you know, ok, this makes sense, what you’re telling me. Or like if somebody uses something and says, “I got really, really itchy,” and then it looks like heroin on the FTIR and you’re like, Oh, that makes sense, ‘cause we know that acetylcodeine, the byproduct of heroin, causes itching and rashes.

I have also seen drug checking services produce this coordination by engaging people who sell drugs in drug checking services. The goal is to equip them to better inform consumers and, in a way, act as a secondary distribution point for the knowledge schema they are helping to produce. In interviews, I’ve heard reports of suppliers choosing to cut the drugs they were selling after learning it contained fentanyl in order to reduce overdose risk or passing along the information they gained through drug checking services to customers to give them a better idea of what their drugs contained. I’ve even heard several stories of suppliers disposing of batches they determined to be unsafe to use—taking a significant financial hit as they did. Brittany recalled one such encounter:

There’s one particular person who probably comes in once or twice a month like “I got this new batch of stuff in.”…They actually had an example of one of the synthetic cannabinoids [in their batch], so they used it and one of their friends used it, and both of them had really severe experiences. They brought it in and described what happened, and we were able to send it to [a bigger lab] for confirmation, and it came back as a synthetic cannabinoid. And they were like, “Yep, flushing the whole thing down the toilet. Don’t give this to anybody else.”

If each of these small acts of care can be considered a form of protection within the criminalized drug market, we can also consider efforts to engage drug checking services as an act of social resistance against that criminalization. They are producing knowledge that criminalization had previously prevented ever being produced. Further, these knowledge-production practices are founded upon the basic premise that people who use drugs are “authorized seers”—reliable and rational individuals who are perfectly capable of looking upon and making sense of the world around them. It should not surprise us that the same “regimes of sight” and technological registers engaged by law enforcement—including claims to experiential knowledge of what fentanyl “feels” like—are also central to the production of knowledge through drug checking. Indeed, Jasanoff has observed that “acts of popular resistance…draw on the same repertoires of technoscientific imagination and instrumental action” as the powers they aim to resist (Jasanoff, 2015). This does not undermine the aims or the impacts of drug checking efforts. But, ironically, in defying the authority of law enforcement as one arm of the state, those wielding and using drug checking technologies are reproducing the drug safety and consumer protection functions that should be carried out by another; they are, indeed, the FDA of Siler Street.

To address the question of how law enforcement institutions have responded to the expansion of drug checking as an overdose prevention strategy, I offer two very different examples of police policy—both extreme but important insofar as they illustrate the impacts of law enforcement’s commitment to different political principles. On one end of the spectrum, Emily’s organization provides mobile services and focuses much of its outreach in an open-air drug market with a heavy police presence. Early in the drug checking program, a deputy superintendent communicated to Emily in no uncertain terms that offering drug checking services would give officers probable cause to arrest her staff. Months later, when policing in this area was stepped up, several staff persons were, in fact, threatened with arrest by local police. Emily was livid:

They didn’t communicate anything to the public health people. And, like, we’re trying to f---ing advocate for our people out front…And then our own police force tells us they’re gonna f---ing arrest staff if we “intervene” with our harm reduction activities? And, yea, like, I’m gonna f---ing take [drug checking] off the street, because, like, I can’t trust our own police force, like, who can very easily f--- me and f--- my f---ing team.

On the other end of the spectrum, police officers serving the area where Brittany and Nathan work have generally embraced drug checking as a harm reduction strategy. The local department has even joined a state-wide initiative in which police officers, themselves, distribute fentanyl test strips to the public. A press release described the effort as follows:

Through this pilot project, fentanyl test strip kits will be distributed by participating police departments and their partners, such as recovery coaches, to develop trust and build relationships with community members struggling with substance use disorders. This engagement tool may help save lives and foster a person’s readiness for a referral to treatment and long-term recovery. (Police Assisted Addiction Recovery Initiative, 2020)

But even this positive outcome raises key questions that should be considered as drug checking services expand in a legally ambiguous environment. By acknowledging the validity of drug checking as overdose prevention, has law enforcement ceded some degree of epistemic authority over the drug market to community partners, or has that authority been retained? Can law enforcement leverage their epistemic authority to tell a different story about drug checking than their harm reduction partners tell, framing the necessary end goal of drug checking as progression toward abstinence-based recovery, for example, rather than simply staying alive? If officers have become somehow “resigned to the fact that the illegal drug use is going to happen,” what are the limits of tolerance for that resignation, and at what point along that continuum of acceptance could this “soft” public health/law enforcement partnership become implicated in other, more punitive police responses to drug use? Drug checking services are poised to expand rapidly across the United States, and, as they do, these questions about how drug market knowledge is produced, who has the authority to produce or validate that knowledge, and which groups or institutions have their authority reified by the knowledge that is put into social or media circulation will be key for understanding how these technologies are shaping the socio-political terrain in which overdose, as a social, practical, and moral problem, is perceived and addressed.

Conclusion

In this article, I have argued that law enforcement actors maintain epistemic authority over the illicit drug market by adopting socially acceptable and naturalized modes of collective seeing, and by presenting in the problem of drugs circulating outside of bodies in a technical register as one that only law enforcement institutions can resolve. These law enforcement-specific “regimes of sight” are mutually constituted with the systems that criminalize and practices that police illicit drug market participants, which directly produce many of the harms we associate with illicit drug use.

I have also argued that drug checking technologies in the hands of people who use drugs, who have been systematically invalidated as sources of knowledge and expertise, produce new and necessary forms of knowledge about the illicit drug market. Through its capacity produce replicable chains of abstractions, drug checking allows for the development of novel and relevant categorical schemas to make sense of variation within the illicit drug market. It further allows for the coordination of those schemas across wide populations of people who use drugs in alignment with personal and phenomenological expertise about illicit drug products that people who use drugs already possess.

Both arguments reveal how technology may be framed as a non-human truth telling device. Further, drug checking demonstrates one way in which technological instruments may have the capacity to lend ontological credibility—and perhaps even some degree of epistemic authority, depending on the audience—to marginalized groups by virtue of the political claim that they are “beyond politics” and perfectly, objectively represent the world. The interaction between that newfound capacity and the extant epistemic authority of law enforcement can take a variety of forms, from volatile antagonism (i.e., police threatening arrest) to tentative stasis (i.e., police enthusiastically joining the ranks of drug checking advocates).

Sheila Jasanoff has observed that institutions, in order to shore up their authority over a problem, seek persuasive ways to talk about it, and are thus always at risk of oversimplifying that problem into seductive but problematic categories: “While institutional discourses often tacitly merge normative and technical repertoires…they may also enable reasoned action by defining the boundary between the promising (‘natural’ or ‘safe’) and the fearsome (‘unnatural’ or ‘unsafe’) aspects of nature and technology” (Jasanoff, 2004, p. 41). The discredited social identity of people who use drugs is one of those problematic categories, which serves the institutions who rely on it yet can be deadly for those individuals to whom it is applied. My hope is that this analysis will prompt us to consider not only how our systems of knowledge are produced and for whom the benefits of that knowledge are intended, but also how mechanisms of social distinction and othering can exclude key experts from participation in knowledge production. The deadly effects of drug criminalization in the United States shows all to clearly the devastating impacts such epistemic injustice can produce.