The Smell of Power

On 8 July 2005, the day after the London bombings, the International Association of Chiefs of Police issued its new guidelines on the detection and prevention of suicide bombings. The IACP is the primary organization through which senior police executives across the globe try to coordinate their powers and practices, and it does this through a number of ‘Training Keys’. The two ‘Training Keys’ issued in July 2005, numbers 581 and 582, were called Suicide Bombers: Part I and Suicide Bombers: Part II. Training Key 582 concerns the use of lethal force, emphasizing shooting at a suspect’s head and thus shooting to kill. Training Key 581 instructs police officers in how to detect potential suicide bombers. It focuses on a set of behavioural and physical characteristics, including sudden changes in behaviour such as a religious person starting to visit sex clubs, the wearing of loose or bulky clothing in the summer, fidgeting or pacing, failure to make eye contact, mumbling, being overly concerned about their bags, strange hair colouring. Among this range of ‘suspicious’ things are also some indicators based on smell: ‘is the individual wearing too much cologne or perfume, or does he or she smell of talcum powder or scented water?’ To reinforce this point, the Training Key references the advice given to border officials by the Department of Homeland Security, to look out for individuals who ‘may smell of herbal or flower water’. Smell is here treated, quite casually, as a technique of police power.

Aside from a few critics noting the racial profiling implicit in the instructions, such comments concerning smell have been largely overlooked. This oversight is symptomatic of a much wider focus within ‘surveillance studies’ on sight and vision. Extensive analyses of CCTV, spy cameras and technologies of global satellite tracking follow a long tradition which has homed in on the ocular as a locus of power. It is the gaze that has dominated critical discussions of surveillance. The ubiquity of the panopticon as a theoretical tool for grappling with the powers of discipline and punishment captures this exactly, but there are a host of other critical analyses which have centred on the practice of ‘seeing like a state’. After sight, it is mechanisms of aural surveillance such as eavesdropping, bugging and phone-tapping which dominate the debate. Little is said about how the state smells, despite the obvious implications that arise when thinking about the powers of sight and sound in the body politic. The ‘new surveillance’ studies has failed to analyse smell as part of the state’s sensory apparatus despite being founded on the ‘extension of the senses’, [1] and the literature on identity as surveillance has picked up on the eye, face, finger, hand or voice in its discussions of the body, but rarely the smell: a major and increasingly influential edited collection such as Jane Caplan and John Torpey’s Documenting Individual Identity, for example, contains just one comment on smell, in an essay on ‘body surveillance’ by one leading surveillance studies thinker, and that is to dismiss the ‘hype’ surrounding any new technology in the field. [2] Why is this? Why, with a critical intellectual culture saturated with analyses of biopolitics, biosecurity, biosurveillance and biometrics, has so little been said about the smell of power? Why is the state’s ‘nosiness’ still understood almost solely through the ocular and the aural?

One reason might be that ‘olfaction is often deemed the least “intellectual”, or the least informative, of our senses’. Seemingly dealing with ‘the airy, the insubstantial, and the formless’, olfaction appears to be a poor cousin to sight when it comes to power. Compared to sight, ‘whose perceptions are so rational that they may be analyzed by the laws of geometry, olfaction rejects geometrical analysis’ in much the same way that it rejects localization. [3] On this view, there is little for the state to be doing with smel . ‘Due to its marginal and repressed status in contemporary Western culture, smel is hardly ever considered as a political vehicle’, notes one cultural history; ‘power appears odourless’. [4]

Yet power’s appearance of being odourless does not mean that smell is not also political. We know well enough that smell clearly has been a political vehicle, from the deodorization of the nineteenth-century city as an exercise in class power through to its articulation as a means of delineating the otherness of the foreign, the state has long exercised a kind of olfactory police as part of its fabrication of order. But really to grasp this we need to understand how the state is now concerned less with merely eradicating the disgusting smell or mobilizing against the foreign smell, and much more with using it. Crucial to this process is the sniffer dog.

Critical theory has never really asked the question:

what is the political meaning of a sniffer dog? As a sniffer dog it has an obvious range of uses, as witnessed by the snarling creatures confronting prisoners being tortured at Abu Ghraib and protestors being ‘kettled’ in London. But as a sniffer dog its meaning lies less in its ability to generate fear and more in its ability to enact a politics of smell; the use of the Chihuahua as a sniffer dog in Japan and the attempt to replace the sniffer dog with sniffer technologies are evidence of the fact that it is the power to smell rather than the power to scare that is important. Investigating this power situates the sniffer dog at the heart of police power. As such, an investigation of the sniffer dog takes us past the banalities of surveillance studies and civil liberties discourse, and straight to the heart of the state’s role in the permanent reinvention of bourgeois order.

It is a little-noted fact about the process of personal identity that before fingerprints were introduced as a primary means of police identification, Barruel had offered to the French police his ‘discovery’ concerning the odour of blood when, in 1829, he proposed ‘smell prints’ as a means of identifying criminals. [5] Not much came of it then, and it would appear to have died out as a result of the growth of fingerprinting in the late nineteenth century and advances in haematology which ruled out blood as the basis of smell prints. Yet regardless of his focus on blood, Barruel was indeed on to something with the idea of smell prints.

In 2007 it was revealed that German police had collected scent traces of activists prior to the G8 summit in Heiligendamm in June of that year. At the time, many thought this an unusual infringement of civil liberties and a new indication of how far the state is now willing to go in intelligence-gathering. But it was neither unusual nor new. In her analysis of ‘Stasiland’, Anna Funder reports visiting one room in the Stasi museum in Leipzig and seeing a cabinet containing glass jars. The woman who ran the museum explained that the jars were ‘smell samples’.

The Stasi had developed a quasi-scientific method, ‘smell sampling’, as a way to find criminals. The theory was that we all have our own identifying odour, which we leave on everything we touch.

These smells can be captured and, with the help of trained sniffer dogs, compared to find a match … The Stasi might break into someone’s apartment and take a piece of clothing worn close to the skin, often underwear. Alternately, a ‘suspect’ would be brought in under some pretext for questioning, and the vinyl seat he or she had sat on would be wiped afterward with a cloth. The pieces of stolen clothing, or the cloth, would then be placed in a sealed jar. The containers looked like jam bottling jars. A label read ‘Name: Herr [Name]. Time: 1 hour. Object: Worker’s Underpants’. [6]

Funder adds that the number of jars suggests that the Leipzig Stasi had smell samples of the entire political opposition in the area. The jars disappeared soon after the 1989 revolution, but then turned up in the office of the Leipzig police, who thought they might still be of use after the revolution and thus kept them in a ‘smell pantry’. So the collection of scent traces of political activists by German police may well be an ‘unsavoury’ reminder that German security agencies are ‘using methods that the Stasi once practiced’, as Hans-Christian Stroebelle, a Greens leader, put it. [7] But what if it can’t be explained away as a relic of the Stasi period in German history? What if something else is at stake, something far more telling about liberal democratic regimes? What if the liberal state is more interested in smell than we realize?In the early 1960s the CIA produced a fairly lengthy report called Human Scent and Its Detection. The document considers in great detail the science of sweat production, including several paragraphs discussing the different functions of the eccrine, apocrine and sebaceous glands, and noting the different ways in which they produce smells. The political purpose of the scientific detail is to raise the possibility of estimating the amount of odour that a human body might produce and the distance that might enable a dog to smell a person. The document suggests that such research might have an important role in future intelligence work. More recently, the Defense Advanced Research Projects Agency (DARPA), the US state’s key agency for military research, has developed a ‘Unique Signature Detection Project’, formerly known as the ‘Odortype Detection Program’, and has been working since 2007 on a system of Identification Based on Scent (IBIS). ‘DARPA wants to be able to detect, track, and even positively identify them [criminals and terrorists] from a distance … using nothing more than the heat and sweat that emanate from a person’s pores’, notes the Information Awareness Office report ‘Detecting Sweaty, Smelly Security Threats’. [8] Both the Department of Homeland Security in the USA and the Ministry of Defence in the UK fund research into smell, in the form of Remote Air Sampling Canine Olfaction (RASCO). [9] Significantly, RASCO is said to be more accurate than, say, a retinal scan, because a person’s smell is thought to be less controllable than their eyes, and dogs are trained to smell past the attempts to camouflage real smells with false ones.

Such projects have been assisted by the chemical industries aiming to get in on the security bandwagon:

the European Network of Excellence in Artificial Olfaction and the International Society for Olfaction and Chemical Sensing now list security politics as a major research interest.

In virtually all cases the focus of the research points to the sniffer dog as key to the politics of smell.

Police forces work on the assumption that police dogs can detect a whole range of substances on a person.

Note that this means that for the most part the dogs are used not to identify individuals, along the lines of a fingerprint or retinal scan, but to identify (or, more correctly, to appear to identify) substances. They are frequently employed in order to ‘catch’ a whiff of drugs on people as they go by and have therefore been a key weapon in the ‘war on drugs’. Yet anyone who travels on public transport will have noticed a rather significant increase in the use of dogs around transport terminals in the last decade. This is often said to be about continuing the fight in the ‘war on drugs’, but it is also clear that the increased presence of police dogs in public spaces is an outcome of more generalized security measures in the ‘war on terror’. ‘Dogs are the public face of the war on terror’, as one journalist puts it. [10] The use of dogs by the state has increased so much that they now have their own publication, the bi-monthly K9 Cop Magazine, launched in 2009.

One of the main functions of the dogs, as the public face of the ‘war on terror’, has been identical to their role in the war on drugs: to justify ‘stop and search’ routines. Dogs sniff the air, the person and their bags, and the police officer reacts accordingly. If it is the case that ‘what smells good is good’ and, conversely, that ‘what smells bad is bad’, [11] then from the police perspective what smells suspicious is suspicious. Yet it is also the case that anything appearing to the police officer to be ‘suspicious behaviour’ is grounds for a suspect to be stopped and searched, including trying to avoid the dogs or even looking like one might be trying to avoid them. Once this occurs the police have ‘reasonable grounds’ to engage in a stop-and-search routine. Thus one can be stopped without knowing whether one is caught up in the war on drugs, war on crime or war on terror at that precise moment in time. The sniffer dog is thus both an emblematic and a symptomatic figure in the universal warfare of contemporary bourgeois order.

The standard civil liberties approach to this question is to point to the fact that sniffer dogs have been found to be wrong in their ‘judgement’ somewhere in as many as three out of five searches, or that the dogs can be influenced by the way their handlers hold their leads, and that the dogs are often overly trusted by handlers who think the dog is never wrong (an assumption reinforced by a wider such belief among the public). [12] But such arguments fail to address the ways in which the dog is, in effect, a technology of state power. A series of case laws across the Western world challenging the use of sniffer dogs reinforces this point. In New South Wales, the Council for Civil Liberties (CCL) challenged the use of police dogs through the case of a person stopped by a plain-clothes police officer who had been working with a dog near a nightclub. As the defendant walked by, the dog sniffed the air, flared its nostrils and then started sniffing the defendant. The police officer’s search of the defendant then found cannabis. The CCL claimed that the use of dogs in such a way amounted to an illegal search, but the NSW Supreme Court held that the use of the dog was not unlawful. The basis of the judgment was that ‘a police officer would have been entitled to walk in the vicinity of the appellant and, if he were to smell cannabis leaf in the appellant’s possession, form a reasonable suspicion sufficient to entitle him to search the appellant.’ The fact that a dog was being used to do this did not change the situation. In other words, as the Supreme Court made clear, it was ‘treating a drug detection dog as an extension of the police officer, an aid to his olfactory senses’. [13] More or less the same legal interpretation applies in all Western societies, as Amber Marks research suggests. [14] Thus, legally, the dog is an extension of the police officer. If the police officer is the long arm of the law, the police dog is the nose of the state.

Note, however, that in contrast to the occasional use of smell by a police officer, dogs are always already sniffing something. In stark contrast to police officers and their far more limited olfactory powers having to deliberately move in close to check a smell, the police dog is permanently sniffing, and permanently sniffing a far wider area than might be sniffed by the human police figure. That is, this extension of police power means that when a dog is in the vicinity, a permanent and sweeping olfactory search of a wide and indefinable space is taking place. The presence of the dog means the state is engaged in a de facto search of an indefinable area and persons in the vicinity, creating the grounds for a de jure stop and search. A critical analysis of the sniffer dog, then, must deal with the nature of ‘stop and search’; and this has to be understood not in terms of surveillance or civil liberties, but in terms of the violence of original accumulation.

suspicion, possession, oppression

Writing about original accumulation in Capital, Marx outlines what he calls the ‘police methods’ used from the early sixteenth century to accelerate the accumulation of capital.

Henry VIII, 1530: Beggars who are old and incapable of working receive a beggar’s licence. On the other hand, whipping and imprisonment for sturdy vagabonds. They are to be tied to the carttail and whipped until the blood streams from their bodies, that they are to swear on oath to go back to their birthplace or to where they have lived the last three years and to ‘put themselves to work’… James I: Anyone wandering about and begging is declared a rogue and a vagabond. Justices of the peace in Petty Sessions are authorized to have them publicly whipped and to imprison them for six months for the first offence.

And on it goes through a series of ‘terroristic laws’ (his term again) perfected through four centuries of capitalist development, through which the masterless men and women of the displaced peasantry were whipped into shape as the working class and bourgeois order was fabricated.

The laws in question were the various Vagrancy Acts passed throughout the Western world. One of the features of such laws was their vague and incredibly broad scope, as they formed an all-encompassing discretionary power which could be used more or less at will by agents of the state. They allowed individuals to be stopped and searched purely on the grounds of suspicion; they were, in effect, the original ‘sus’ laws.

The Vagrancy Act of 1824, for example, provided for ‘the Punishment of idle and disorderly Persons, and Rogues and Vagabonds’. It was this Act which constituted the core police power exercised by the ‘new’ British police created in 1829 (and which explains why the Act creating the ‘new police’ was so thin: because the powers that were to be exercised by the new police already existed in the Vagrancy Act). The powers in question were consolidated with the Metropolitan Police Act 1839, which allowed the police to stop and search ‘any Person who may be reasonably suspected of having or conveying in any Manner anything stolen or unlawfully obtained’.

The ambiguity built into the laws and the breadth of powers they offered means that they were ‘sweep laws’: they allowed individuals to be stopped on the grounds merely of suspicion, and even perhaps on the grounds of suspected criminal intent. Hence a huge range of acts or non-acts, forms of behaviour and inactivity, fell foul of the laws, such that a ‘sweep’ of any area would easily find guilty parties. In granting sweeping powers to target anything considered ‘disorderly’ or ‘suspicious’, the determining principle underpinning the vagrancy laws was not the liberty of the subject but the well-ordered society. Once we grasp that policing is about the fabrication of social order rather than ‘law enforcement’, the vagrancy laws can be seen as the police power par excellence; which is to say, they were a fundamental weapon in the class war.

Much as the heyday of vagrancy laws was the sixteenth through to the nineteenth century, the laws themselves remained in place until the second half of the twentieth century, at which point their vagueness, their criminalization of status without any particular criminal act being committed, and their overt class and race bias, came to be seen as problematic. Symptomatically, however, their abolition coincides with the birth of the quintessential social wars of the late twentieth century: the wars on drugs and crime. The key operative law in these wars has been the law of possession. In such laws, possession qua possession becomes the crime itself, or can be used as evidence of intent to commit a crime. And so the law on possession has become what Markus Dubber calls the sweep offence of choice in modern policing, being used in exactly the same manner as the original vagrancy laws. [15]

In offering sweeping powers, ‘possession’ shares the central advantages of ‘vagrancy’ as a policing tool: it is flexible and convenient, it relies heavily on the idea of suspicion, it panders to the law’s belief that intent can be identified, and it continues the state’s declared mandate to determine what constitutes a public threat and eliminate it accordingly. Moreover, in being focused on possession rather than vagrancy, possession law not only expands the scope of police powers into the home, thereby exploding the liberal myth of privacy, but also functions in a way that means millions of people commit some form of crime every day or could be suspected of being about to commit a crime. They often don’t know this, until stopped and searched by a police officer after having been ‘detected’ as possibly possessing something suspicious. The list of items for which people have been found in possession is almost endless: drugs, drug paraphernalia, firearms and other weapons are obvious candidates for such a list. But it would also include imitation weapons, toy weapons, air pistols and rifles, any implement or instrument that might be used as a weapon, ammunition, body vests, equipment that might be used in burglary, stolen property, instruments that might be used for graffiti, counterfeit trademarks, unauthorized recordings of performances, public benefit cards, forgery devices, embossing machines, vehicle identification numbers, prison contraband, obscene material, eavesdropping devices, fireworks, noxious materials, slugs, and on it goes, right down to items deemed by the police to constitute a threat to ‘security’. This list of items, compiled by Dubber from possession cases in just one US state, New York, could be extended to include those people held for possessing spearfishing equipment (in Florida) and undersized catfish (in Louisiana).

Now, the extent to which this police power permeates the social world can be seen in the fact that many possession offences begin with a simple traffic stop, a salutary reminder that traffic police are never just policing traffic. But it can also be seen in the fact that many possession offences start from the nose of a sniffer dog. The main law brought into action when sniffer dogs are thought to detect anything is the law on possession. Because of the more or less permanent ‘sweep’ mode that comes from the dogs sniffing everything and anything that passes by them or even near them, the ‘sniff’ very easily generates ‘reasonable grounds’ for an officer to stop and search; the sniff produces the suspect. The mythical reliability of the dog means that the initial sniff is virtually unchallengeable in the courts and thus the police officer has very simple and universal grounds for stopping and searching.

Note, however, that most uses of stop-and-search powers actually reveal very little ‘crime’, in the sense of leading to formal arrests and prosecutions. What stop-and-search powers do, rather, is open the possibility for the police to impose their authority as state officials on individuals and groups considered problematic, threatening, or, in the terms of that key concept of police power, ‘disorderly’. The sniffer dog plays a crucial role in this process, thereby drastically extending the sensory reach of the technology of the state.

This is what explains the increased presence of dogs in public spaces since 2001, but also throughout the disciplinary apparatus of the modern state, in institutions such as schools, prisons and mental health institutions (visits to which are sold to us as exercises in ‘community policing’ but which are designed to nurture greater social acceptance of the dogs and to carry out searches – de facto and thus de jure – while on the premises). The inflection of the war on vagrancy (that is, the class war) into the war on drugs, the war on drugs into the war on crime, and the war on crime into the war on terror, and then the war on terror back into the wars on drugs and crime, is a mechanism for articulating the enemy as existing anywhere and everywhere in a way that requires police infiltration of all space. The power of smell plays a crucial role in this infiltration. In the endless warfare of contemporary political order in which the complicated cultural practices surrounding drugs, the social dynamics surrounding crime and the political tactics surrounding terrorism are reduced to an amorphous and ubiquitous ‘enemy’ of good order, police discretion becomes a key to victory. And in this war, the sniffer dog is in the front line.

Yet one of the crucial features of this war is precisely that there is no front line in the traditional military sense of the term. This is why so much effort is being made to mechanize the sniffer dog. Much is made in the media of such new technologies: from the CIA’s original attempt to create a ‘mechanical dog’ (which, ‘when he is born’, the CIA report notes, ‘should be much more unobtrusive than his natural ancestor, should be able to tell us just whom he has smelled, and should maintain a reliable permanent record of his visitors’); DARPA’s spending in excess of $3 million a year since 2001 developing an E-Nose; the Commonwealth Scientific and Industrial Research Organization’s ‘cybernose project’ (tagline: ‘sniffing a better future with Cybernose’), which prioritizes cheese, wine and biosecurity; the new flashlight called the Passive Alcohol Sensor, which contains a new smell detection technology (‘the sniffer’, as the police call it, ‘an extension of the police officer’s nose’); right down to a new sensor called Scentinel to identity individual body odours, developed by a UK company called Mastiff Electronic Systems (note the dog reference in the name). Many of these developments imagine the technological sniffer as electronic filing system as well as the detector of the initial information, and reconnect the politics of olfaction back to the question of identification. But, as I have been arguing, the real political issue is not the smell of identification; it is how the smell of power functions in terms of the fabrication of order.

Thus far more telling is the attempt to develop technology that reinforces the obliteration of any front line in the new universal war: the ‘sniffer helicopter’.

By radically extending the potential area to be sniffed, this technology makes possible a significant shift in the policing of territory, extending the powers of olfaction from individual officers and their dogs in a relatively small amount of space to a much wider ground. If successful, such mechanization will overcome the limitations imposed on the state’s use of real dogs – the training, maintenance and handling are expensive, time-consuming and cumbersome – and will be another tool in the technological armoury of the state.

This move, from ‘canine cop’ to ‘techno cop’, will be one more step towards realizing the state’s dream of exercising a permanent and ubiquitous police power to obliterate anything regarded as disorderly.


1. ^ Gary T. Marx, ‘What’s New about the “New Surveillance”: Classifying for Change and Continuity’, Surveillance and Society, vol. 1, no. 1, 2002, pp. 9–29; p. 16.

2. ^ David Lyon, ‘Under My Skin: From Identification Papers to Body Surveillance’, in Jane Caplan and John Torpey, eds, Documenting Individual Identity: State Practices in the Modern World, Princeton University Press, Princeton NJ, 2001, p. 300.

3. ^ F. Gonzalez-Crussi, The Five Senses, Vintage, New York, 1991, pp. 65, 71.

4. ^ Constance Classen, David Howes and Anthony Synnott, Aroma: The Cultural History of Smel , Routledge,

London, 1994, p. 161.

5. ^ Alain Corbin, The Foul and the Fragrant, Papermac,

London, 1996, p. 187; Simon A. Cole, Suspect Identities: A History of Fingerprinting and Criminal Identification, Harvard University Press, Cambridge MA, 2002, p. 12.

6. ^ Anna Funder, Stasiland: Stories from Behind the Berlin Wal , Granta, London, 2003, p. 8.

7. ^ Cited in Erik Kirschbaum, ‘Germans Outraged by “Scent Profiling” Ahead of G8’, Reuters, 23 May 2007.

8. ^ ‘Detecting Sweaty, Smelly Security Threats’, 27 April, 2010, [archive]

9. ^ Spencer Tebrich, Human Scent and Its Detection, CIA Historical Review Program (n.d., originally for official use only, released 22 September 1993 – the release date suggests that the document was produced in 1963); Department of Homeland Security, Five Year Research and Development Plan, August 2008; ‘Second Memorandum from the Ministry of Defence’, to the House of Commons Select Committee on Defence, 29 September 2006.

10. ^ John Hind, ‘You Are Being Sniffed: Why Dogs Are the Public Face of the War on Terror’, Observer, 5 March 2006.

11. ^ Anthony Synnott, The Body Social: Symbolism, Self, and Society, Routledge, London, 1993, p. 190.

12. ^ See, for example, New South Wales Council for Civil Liberties, Review of the Police Powers (Drugs Detection Dogs) Act, August 2004, p. 4; David Grainger, ‘Sit! Stay!

Testify!’, Fortune Magazine, vol. 149, no. 2, 26 January 2004.

13. ^ Darby v. Director of Public Prosecutions, NSWCA 431 (26 November 2004), para.

62. ^

14. ^ Amber Marks, ‘Drug Detection Dogs and the Growth of Olfactory Surveillance: Beyond the Rule of Law?’, Surveil ance and Society, vol. 4, no. 3, 2007, pp. 257–71.

15. ^ Markus Dirk Dubber, ‘Policing Possession: The War on Crime and the End of Criminal Law’, Journal of Criminal Law and Criminology, vol. 91, no. 4, 2001, pp. 829–996.