Réhabiliter l'odorat
comme un sens aussi noble qu'essentiel à la pensée.
Les philosophes ont
généralement fait peu de cas de l'odorat, tenu pour le moins sûr
et le moins noble de tous les sens, et ont préféré concevoir la
connaissance sur le modèle de la vue ou de l'ouïe, comme s'il
allait de soi que la vérité se voit ou se contemple, mais qu'elle
ne se sent pas. Ainsi exclu du domaine de la connaissance,
l'odorat s'est aussi vu régulièrement refuser toute légitimité en
esthétique. Bref, l'idée prévaut assez largement que ni le vrai
ni le beau ne sauraient être l'affaire du nez.
La physiologie et
l'anthropologie nous apprennent que le nez est infiniment plus subtil
qu'on ne le dit et joue un rôle majeur dans nos existences
individuelles et sociales. Condillac imagine une statue qui, à
l'origine privée de tout sens, développe peu à peu les différentes
facultés à partir du seul odorat. Nietzsche - "Tout mon génie
est dans mes narines" - fait explicitement du "flair"
une vertu philosophique et un instrument de connaissance à même
d'atteindre un réel inaccessible aux autres sens.
Comment les chiens "voient-ils" avec leur nez
Non seulement nous ne sommes pas toujours en train d'utiliser notre odorat, explique Alexandra Horowitz, mais lorsque nous l'utilisons, c'est souvent parce qu'on sent une bonne ou une mauvaise odeur: il s'agit rarement d'une simple source d'information. Les odeurs que nous percevons la plupart du temps sont soit agréables soit répugnantes, très peu ont un caractère aussi neutre que notre vision... de même que nous voyons le monde, le chien le sent.
Il y a deux conduits aériens dans la truffe
(entre 2.000 et 3.000 millions de capteurs, contre 6 millions chez
l'homme) : l'un pour respirer, l'autre pour sentir le monde environnant, percevoir les mouvements, ou encore faire la
différence entre amitié et hostilité. Plus impressionnant encore, le
nez du chien l'aide à voir des choses qui ne sont plus ou pas encore
visibles: il détecte ce qui a eu lieu plus tôt (la voiture garée à
un emplacement il y a quelques minutes) et ce qui n'est pas encore
arrivé (une personne qui va apparaître au coin de la rue).
Nous sommes peut-être capables de remarquer que quelqu'un a ajouté une cuillère de sucre à notre café, mais le chien peut détecter cette cuillère de sucre dans l'équivalent de deux piscines olympiques.
Mais l'être humain aussi a du flair !
Très instructif "De la molécule à l'odeur".
Forensic Science International 209 (2011) 133–142, Lauryn E. DeGreeff, Allison M. Curran, Kenneth G. Furton
Abstract
Human scent can be
collected by either contact or non-contact sampling mode. The most
frequently used human scent evidence collection device known as the
Scent Transfer Unit (STU-100) is a dynamic sampling device and is
often used in a non-contact mode. A customized human scent collection
chamber was utilized in combination with controlled odor mimic
permeation systems containing five standard human scent volatiles to
optimize the flow rate, collection material and geometry of the
absorbent material. The scent collection method which yielded the
greatest amount of volatile organic compounds (VOCs) detected
included the use of a single layer of Johnson and Johnson
gauze/multiple layers of Dukal gauze with the STU-100 on the lowest
flow rate setting. The correlation of the resulting VOC profiles
demonstrate that collection of standard VOCs in controlled conditions
yielded reproducible VOC profiles on all materials studied with the
exception of polyester. Finally, the method was tested using actual
human subjects under optimized set of conditions.
Collection and identification of human remains volatiles by non-contact, dynamic airflow sampling and SPME-GC/MS using various sorbent materials
Lauryn E. De-Greeff et Kenneth G. Furton - , in Anal Bioanal Chem (2011) 401:1295–1307Collection and identification of human remains volatiles by non-contact, dynamic airflow sampling and SPME-GC/MS using various sorbent materials
Individual human scent as a forensic identifier using mantrailing
- Forensic Science International, Vol 282, January 2018, pp. 111-121 LeifWoidtke, JanDreßler, CarstenBabian
Decomposing Human Blood: Canine Detection Odor Signature and Volatile Organic Compounds
Rendine M, et al. J Forensic Sci. 2019.
Abstract
The
admissibility of human "odor mortis" discrimination in courts depends
on the lack of comprehension of volatile organic compounds (VOCs) during
the human decay process and of the lack in standardized procedures in
training cadaver dogs. Blood was collected from four young people who
died from traffic accidents and analyzed using HS-SPME/GC-MS at
different decompositional stages. Two dogs, professionally trained, were
tested to exactly locate blood samples, for each time point of the
experiment. We found a long list of VOCs which varied from fresh to
decomposed blood samples, showing differences in specific compounds. Dog
performance showed a positive predictive value between 98.96% and 100%
for DOG A, and between 99.47% and 100% for DOG B. Our findings
demonstrated that decomposing human blood is a good source of VOCs and a
good target for canine training.
Cadaver dogs: Unscientific myth or reliable biological devices?
IreneRiezzo,MargheritaNeri,MarcelloRendine,AlessandroBellifemina,
SantinaCantatore, CarmelaFiore, EmanuelaTurillazzi
We aimed to detect the
reliability of dogs trained to locate human cadaveric blood.
We adopted an optimized
and rigorously controlled experimental design.
The primary detection
task was human cadaveric blood at very low concentrations.
Dogs’ discriminative
capability among confounding substances was also investigated.
Dogs represent a
scientifically unassailable tool in detecting human blood traces.
Abstract :
Dogs are commonly used to
detect explosives, narcotics, and other illegal materials. In the
forensic setting, cadaver dogs are trained to detect and locate
concealed human remains or fluids due to the high sensitivity and
selectivity of the canine olfactory system and the relative ease with
which dogs can be trained and handled. The need for international and
scientifically validated standards has long been outlined by the
literature. It is important, therefore, to establish the reliability
of the handler/dog team. Our study aimed to detect the real
effectiveness of dogs trained to locate human cadaveric blood in very
low concentrations, through an optimized and rigorously controlled
design which would rule out any possible sources of bias. The study
was designed to determine the dogs’ olfactory sensitivity to human
cadaveric blood and how this capacity might change as the dilution of
blood increases from pure blood to very low concentrations. The
further step was to examine the dogs’ ability to discriminate among
target (human cadaveric blood) and non-target (confounding
substances) odors (discriminative capability). Our results revealed
that well trained dogs were able to detect human cadaveric blood
samples even when very low concentrations of blood were stored in the
tubes, showing high levels of olfactory sensitivity and to
discriminate the target odor even when the non-target odor was orders
of magnitude higher in concentrations. Although our results are based
only on two dogs, the procedure we used may provide a comprehensive
answer to the need for a scientifically unassailable tool for
quantifying and objectifying the performance of well-trained specific
search dogs in detecting human cadaveric blood traces.
09 May 2013
Mary Cablk - The Science
of Sniffer Dogs
Blind and double blind
training can be very simple, and can be made complex. Let's start
with simple. The basics of either (blind or double blind) are the
same:
someone besides the
handler sets the problem.
anyone who has knowledge
about the set (how many hides, where they are) keeps that information
to themselves.
Here's where the
difference comes in: if it is a blind problem, someone with knowledge
about the solution to the problem may be present and they provide no
information to the handler that might bias the outcome. They may
stand on the sideline, they may walk along with the team, they may be
anywhere nearby. In a double-blind problem all individuals with
knowledge of the solution to the set, in whole or in part, are out of
the area. In double blind there isn't a way for the handler to
recieve any information that would help them solve the problem.
There's no ability for bias to occur.
This requires no more
people to set up than is needed for any other training, with the
exception of when a handler is training by themselves without
assistance. A handler cannot self set a blind or double blind
problem. Two people can do blind and double blind training. It can be
that simple and easy.
When do blind and double
blind training problems make sense in training? At some point a dog
and handler have to make the leap from working known problems, where
the dog is learning it's target odor for example, learning it's
trained alert or final response, or figuring out how to follow the
track of the subject through turns and over different surfaces. You
don't want that leap to be made on a real search.
One approach is to begin
with blind problems where a person with knowledge of the solution can
step in if needed. Not truly blind when assistance or feedback is
given, it is a transitional step. Generally this is done in the early
stages of moving away from known problems so that the dog is
continued to be rewarded only on its target odors. At some point the
handler has to go for it, and work problems without any help. In the
words of my colleague Terje Groth Berntsen, Head of the Norwegian
People's Aid Global Training Cenre (http://www.npa-gtc.org/), "Mary!
You have to let the dog fail! How else is it going to learn?!"
Blind problems show where
weaknesses are in the team, such as dependencies, and can be used
effectively to remedy training issues. The challenge in blind
problems is for the knowledgable person to remain neutral. Even when
a knowledgable person tries their hardest to remain neutral, it is
still possible for them to bias the team inadvertantly. This is why
double blind training is necessary to document a team's reliability.
In double blind problems
there is no safety net for the team. Because the handler does not
know the solution, and there's nobody there who can step in to offer
assistance, the dog can leave sources without the handler knowing it,
the dog can alert on non-targets, and the dog can be rewarded for
false alerting or not rewarded for finds. On the other hand, with
nobody around to offer input, the team can show they have proper
training - making independent finds, few to no misses, and few if any
false alerts. For these reasons double blind problems are highly
useful diagnostics of the reliability of a team.
Handlers who work double
blind problems and are able to address training issues discovered in
doing so, report greater confidence and with good reason - they
demonstrate over and over that their dog works independently at what
it is trained to do.
One last practical tip on
blind and double blind: how is the team assessed? The handler marks
alerts in the field with flags, or records their track on a GPS. When
the dog is back up in the truck, the handler and problem setter walk
the area and count coincident flags. For a track the handler's GPS is
downloaded and mapped against the tracklayer's GPS. They can walk the
tracklayer's route together as well.
There are more subtleties
that go along with all of this, whether known, blind or double blind,
and those are for future blogs; what about pool alerts? how far off
track can a dog be and still be "on track"? when is a false
a false?
My data show that when
handlers start working double blind problems their reliablity drops
dramatically at first - and then it increases steadily. Be ready to
see this happen when you first start working double blinds. With time
and continued training using known, blind, and double blinds wisely,
the reliability of the team remains high. This is where all canine
handlers want to be, and can be. It is that simple.
Témoignant pour la défense dans l'affaire Redwine,
Mary Cablk, une chercheuse basée au Nevada qui se concentre sur la détection sensorielle et est un maître-chien de chiens cadavres, a déclaré que les chiens renifleurs devaient être formés à un environnement spécifique et, dans le cas des chiens cadavres, à un niveau de décomposition spécifique. Un chien dressé avec des restes humains frais dans un climat humide ne serait pas fiable pour rechercher des restes humains en décomposition dans une région aride, a déclaré Cablk.
Les chiens cadavres ont besoin d’au moins 16 heures d’entraînement par mois, a déclaré Cablk. Corcoran a témoigné qu'elle entraînait Molly huit heures par semaine. Cablk a également déclaré que les chiens ne devraient pas être entraînés à réagir à l'odeur des cheveux; Corcoran a témoigné que Molly avait été formée pour détecter les cheveux.
Cablk a déclaré que les chiens qui s'entraînent avec des restes humains frais absorbent des substances chimiques différentes de celles des chiens entraînés à détecter des parties du corps en décomposition. Corcoran a témoigné que Molly ne s'était jamais entraînée avec des restes humains ou l'odeur de restes humains en décomposition depuis au moins neuf mois.
Un chien bien entraîné ne trouvera pas seulement une odeur mais localisera le point d'où elle émane. Une odeur résiduelle ne peut rester dans un endroit plus d'une semaine, dit-elle. Elle ne peut pas être quantifiée, et la lumière du soleil et la chaleur peuvent modifier la composition de l’odeur, a-t-elle déclaré.
«Elle ne dure pas éternellement», a déclaré Cablk à propos de l’odeur de restes humains.
Mary Cablk, une chercheuse basée au Nevada qui se concentre sur la détection sensorielle et est un maître-chien de chiens cadavres, a déclaré que les chiens renifleurs devaient être formés à un environnement spécifique et, dans le cas des chiens cadavres, à un niveau de décomposition spécifique. Un chien dressé avec des restes humains frais dans un climat humide ne serait pas fiable pour rechercher des restes humains en décomposition dans une région aride, a déclaré Cablk.
Les chiens cadavres ont besoin d’au moins 16 heures d’entraînement par mois, a déclaré Cablk. Corcoran a témoigné qu'elle entraînait Molly huit heures par semaine. Cablk a également déclaré que les chiens ne devraient pas être entraînés à réagir à l'odeur des cheveux; Corcoran a témoigné que Molly avait été formée pour détecter les cheveux.
Cablk a déclaré que les chiens qui s'entraînent avec des restes humains frais absorbent des substances chimiques différentes de celles des chiens entraînés à détecter des parties du corps en décomposition. Corcoran a témoigné que Molly ne s'était jamais entraînée avec des restes humains ou l'odeur de restes humains en décomposition depuis au moins neuf mois.
Un chien bien entraîné ne trouvera pas seulement une odeur mais localisera le point d'où elle émane. Une odeur résiduelle ne peut rester dans un endroit plus d'une semaine, dit-elle. Elle ne peut pas être quantifiée, et la lumière du soleil et la chaleur peuvent modifier la composition de l’odeur, a-t-elle déclaré.
«Elle ne dure pas éternellement», a déclaré Cablk à propos de l’odeur de restes humains.
Time-dependent VOC-profile of decomposed human and animal remains in laboratory environment
E.RosierS.LoixW.DevelterW.Van de VoordeJ.TytgatE.Cuyper
Human and animal remains
were analyzed after 9 and 12 months of decomposition.
Abstract
A validated method using a thermal desorber combined with a gas chromatograph coupled to a mass spectrometer was used to identify the volatile organic compounds released in decomposed human and animal remains after 9 and 12 months in glass jars in a laboratory environment. This is a follow-up study on a previous report where the first 6 months of decomposition of 6 human and 26 animal remains was investigated. In the first report, out of 452 identified compounds, a combination of 8 compounds was proposed as human and pig specific. The goal of the current study was to investigate if these 8 compounds were still released after 9 and 12 months. The next results were noticed: 287 compounds were identified; only 9 new compounds were detected and 173 were no longer seen. Sulfur-containing compounds were less prevalent as compared to the first month of decomposition. The appearance of nitrogen-containing compounds and alcohols was increasingly evident during the first 6 months, and the same trend was seen in the following 6 months. Esters became less important after 6 months. From the proposed human and pig specific compounds, diethyl disulfide was only detected during the first months of decomposition. Interestingly, the 4 proposed human and pig specific esters, as well as pyridine, 3-methylthio-1-propanol and methyl(methylthio)ethyl disulfide were still present after 9 and 12 months of decomposition. This means that these 7 human and pig specific markers can be used in the development of training aids for cadaver dogs during the whole decomposition process. Diethyl disulfide can be used in training aids for the first month of decomposition.After we die, our rotting bodies release specific types of gases into the environment. When a body goes missing, dogs can sniff for these gases to track it down. But forensic scientists are still trying to nail down the exact “smell of death” that is unique to humans. Now, a team in Belgium has identified seven compounds that only pigs and people produce late into decomposition.
When a body first begins to decompose, enzymes dissolve the cells from within. Next comes bloating and putrefaction, giving the corpse a greenish tinge. Active decay starts when the skin breaks and the corpse begins to liquefy. Soon, only shreds of flesh cling to the body. By the end, only hair and bones are left. Figuring out which chemicals our bodies release at these different stages would also allow cadaver dogs to be trained to find bodies at specific points in their decay.
In the new study, scientists examined a menagerie of remains from six humans and 26 animals that included pigs and other mammals, fish, frogs, turtles, and various birds. The team sampled the gases surrounding the remains at nine and 12 months of decomposition.
Previously, the scientists had investigated the corpses at six months of decomposition, and came up with eight compounds only released by human and pig bodies. The new work confirmed that seven of them are still hanging around later in decomposition.
"This means that these 7 human and pig specific markers can be used in the development of training aids for cadaver dogs during the whole decomposition process," wrote the authors, who published the findings on June 2 in the journal Forensic Science International. The other compound, diethyl disulfide, could be used to detect bodies at the beginning of decomposition.
When the team sampled bodies after six months, the bodies had already passed into advanced decay (the bits of flesh stage). After 12 months, the corpses had not yet dried out, so it’s not yet certain whether bodies still release these compounds very late in decomposition. The remains also stayed tucked in glass jars, so scientists will have to investigate whether changes in moisture or temperature that occur out in the open influence which chemicals are released.
Currently, cadaver dogs are trained on the stench of compounds such as cadaverine and putrescine. But dogs sometimes fail to pick up on real body parts, indicating that these training scents do not capture the full complexity of the real decomposition odor.
"A human specific marker can be used to train cadaver dogs more efficiently and therefore win time to locate a body," the authors wrote in their previous report. "Moreover, when a human specific marker is found, it might be possible to develop a portable device that is sensitive enough to locate human remains."
Cadaver dog and handler team capabilities in the recovery of buried human remains
Alanna E Lasseter, Keith Jacobi, Ricky Farley, Lee Hensel
Abstract
The
detection of human remains that have been deliberately buried to escape
detection is a problem for law enforcement. Sometimes the cadaver dog
and handler teams are successful, while other times law enforcement and
cadaver dog teams are frustrated in their search. Five field trials
tested the ability of four cadaver dog and handler teams to detect
buried human remains. Human and animal remains were buried in various
forested areas during the summer months near Tuscaloosa, Alabama. The
remains ranged in decomposition from fresh to skeletonized. Cadaver dogs
detected with varying success: buried human remains at different stages
of decomposition, buried human remains at different depths, and buried
decomposed human and animal remains. The results from these trials
showed that some cadaver dogs were able to locate skeletonized remains
buried at a significant depth. Fresh and skeletonized remains were found
equally by the cadaver dogs along with some caveats. Dog handlers
affected the reliability of the cadaver dog results. Observations and
videotape of the cadaver dogs during field trials showed that they were
reliable in finding buried human remains.