Increasing the success rate of scent discrimination

Considerations to the Process of Increasing the Success Rate of Scent Discriminating Hounds The hound’s ability to trail is not a matter of nature vs. nurture or genetics vs. training it’s a dynamic process of the contributing elements co-evolving from moment to moment. For example: multicomponent determinants such as, levels of health and fitness, the strength of the hounds ability and desire to follow his nose, his desire to learn, the shared task environments (the handlers abilities, scenting conditions, or terrain), and the way in which circumstances have been put together to facilitate coherence between the hound’s understanding, ability, and the environment.


Learning and development involves the assembly of many components. Biology and experience interact to construct behavior. Neither should be trivialized in favor of the other.

Water, in flowing, hollows out for itself a channel, which grows broader and deeper; and, after having ceased to flow, it resumes, when it flows again, the path traced by itself before.

M. L©on Dumont

Early availability of olfactory discriminating experiences accompanies greater olfactory neural organization within the hound’s sensory system. The fetal and newborn brain starts out with an overproduction of sensory perceptually related neurons that subsequently decline. During development, approximately two thirds of nerve cells (neurons) in the brain atrophy and die that have failed to connect with their neighbors through neural activity. There is considerable developmental evidence that experience plays a critical role in determining which emergent groupings of neurons will survive to form the neural architecture (Doubell, Stewart, 1993; Gottlieb, 1991; Greenough, Black, and Wallace, 1987; McCabe, Horn, 1991; Merzenich, Allard, and Jenkins, 1990; Rosenzweig, Bennett, Diamond, 1972; Sur, Pallas, and Roe, 1990). Also, a lack of early experiences can result in arrested sensory development: the overall shapes of the neurons appear to be genetically determined but the branching of the neural arbors depends on input. It is through the experiential synapse formation that value-based circuitry is constructed. Neural specificity is not preprogrammed: genetics provides only general parameters. Freeman found that for every discriminated odor every neuron in the olfactory bulb participates. The perceptual information exists in the synaptic patterns of neural activity, [developed through experience], not in a small number of neurons unique for each odor (Freeman, 1975, 1987, 1991). Because of the hound’s experiences certain patterns of cell connections are assembled, that are further strengthened during the selection and attendance to particular odorants. Experiences continually open up opportunities for growth, perception and action. As actions are performed they become part of a history that has been continually evolving and changing which contributes to the formation of future action. Meaningful events in the hound’s past can be related to new events. This relationship is established through the affected value system of the individual hound. The ability to assemble and reorganize meaning and behavior in response to changing conditions is the genesis for new adaptive behavior (Kelso, 1995; Thelen, Smith, 1994). The sequence of training experiences i.e. his previous experiences and the subsequent modifications in response to added experiences effects future development.

The hound’s behavior is assembled by the nature of his desires or intentions, opportunistically recruiting the necessary and available internal and environmental components. The desire orients attention to desire related opportunities, which in turn influences intentions to act. What the hound learns is in part mediated by the strength and direction of his desires, combined with environmental conditions and what has worked within his developed repertoire to achieve his goals. The things that are highly valued are attended to more than things that maintain little or no value. They attend to and learn according to what they perceive as being relevant to achieving what they aspire toward. Furthermore, the strength of the hound’s response to a scent depends not just on the stimulus situation, but also, on the hound’s state of attention, his motivational state, and his experiential history. Behavior and learning varies with the state of behavioral arousal. Learning is more likely to take place when perception processes are in an active state. A large part of the hound’s performance and learning is the result of his attentional capacity and the proportion of that capacity that is devoted to the task. Individual differences in attentional capacities can be related to differences in experiences and what, or the amount that has been learned from prior activities. Also attention is not the same as motivation. Attention is prompted by motivation, and the interaction between attention and motivation influence each other. Together they act to modulate the hound’s responsiveness to the environment: too much motivation can override attention. Typically, heightened attentiveness facilitates the development of skills whereas too high or low motivation can disrupt the quality of development and performance. Attention and motivation are directional; they can not be directed at all things at one time. Attention and motivation carry weight or value resulting from positive or negative experiences. Weights change as a result, or lack of experience. What is being learned has to do with where attention is directed. Plus, there is competition among motivational states such as; rest vs. play vs. hunger or thirst vs. enthusiasm vs. security vs. energy levels vs. heath and fitness levels, etc¦ Each motivational state inclines a bias that values one activity over another, given the hound’s current state, the environmental situation, and his prior experiences. What makes the difference between behavioral choices and learning is the strength and direction of his attention, which can be, in part, a factor of what he has learned from prior experiences and his capacity to recognize that there is something to choose from. Alternatives are rarely of equal strength; in most cases one alternative is preferred over others. Thought and behavior is influenced by the hound’s internal state, environmental factors, and his prior perspectives. The level and direction of attention and motivation is an important component in the process of skill development. An important goal will continue to be pursued, whereas a trivial goal will be fraught with distractions along the way. A weak response to a weak stimulus predicts a low probability of a desired result.

The ability to discriminate, developed through an experiential history, separates the relevant from the irrelevant sensory signals in an environment containing a multitude of sensory stimuli. How accurately the hound works depends on his current motivational state and the amount of opportunities detected in the environment that are relevant to the task. The threshold for the most vapid scent is not measured as much in days as it is by transforming environmental conditions and by what the hound has learned to recognize as different, or separate out. The information may be available but the hound may not process it or know to act on it because he lacks discriminative awareness. In this view, the things that are discriminated are as or of more relevance than the limits of sensory transduction. A stimulus must represent something for it to have a meaningful effect and be associated with something else. To respond accurately, timely, and adaptively, in different situations demands a discriminative real-time experiential history from different contexts. Different contexts contribute to clarify the relevant properties of the task. What appears to be a lack of intelligence or olfactory ability can in fact be a lack of discriminative awareness, stemming from unclear training methods, misdirected attention or low motivation which affects the number of things that will be perceived as relevant and remembered. The accumulated effect equates to stinted awareness for the construction of higher order recognition to the task components. The more accurately the hound is able to recognize the relevant stimuli in a working environment, the more advantageous it becomes for him to remember them. The amount of information observed during purposeful behavior can convert slow learning to insightful learning because the amount observed will be greater. Therefore, the probability that prior learning experiences will be observed as corresponding or not with new experiences will be greater because the amount that has been observed is greater, which provides a basis for learning more arbitrary relationships. The depth of prior learning experiences increases the rate of learning. The things that are of high priority are learned more in-depth and are compounded over time measurably more than the things that are of little felt relevance. Intelligence is the ability to adaptively combine and adjust former perspective constructs with changing contexts. It develops and transforms from the background of prior experience and capacities. The more discriminative knowledge the hound has the more awareness will proliferate from fluctuating environments and experiences.

When several stimulus cues are predictive to achieving a desired result individuals may differ in their attention to the different stimulus information and consequently understand the task requirements differently. For example: when following a trail there may be a number of different scents available to the hound that could represent to him the components of the task. Such as a multitude of scents from a well traveled corridor, scent from disturbed ground or crushed vegetation resulting from the subjects traverse, or the scent of the target accompanied with a host of less elusive scents like perfumes or tobacco products. While working the hound will opportunistically attend to the scents that are the easiest to recognize as long as they remain relevant to what he perceives the task requirements to be (read Williams, Johnston, Cicoria, Paletz, Waggoner, Edge, Hollowell, 1998). Conditioning is faster when scents are more salient or discernable. This can present a problem if the hound incorrectly understands the relevance of the primary component to the task. For example: If following crushed vegetation or disturbed ground surfaces predicts its consequence (a reward) and if following crushed vegetation or disturbed ground surfaces is, as I believe, easier to discriminate than the subjects scent, thus making following the subjects scent more economical to follow, then the governing influence of the crushed vegetation or disturbed ground surfaces will be greater than the subjects scent provided they remains predictive of their consequences. Animals will learn to attend more to stimuli that are more economically predictive of their consequences and will come to ignore the less economical stimuli provided their predictive value remains irrelevant. Response to a stimulus can vary considerably from one hound to another depending on the amount of attention, exposure, and validity the hound attributes to a stimulus. Other concomitantly present stimuli, cues, or features of the environment compete with the subjects scent for association to the relevance of the task or attention. The differences in the understood validities of the stimulus components to the trail can result in a significant domination of attention to one component over another or to a broadly generalized understanding of the task. Individual differences of attentional dominance to a particular stimulus in the environment will result in different solutions to a given task. The more the hound attends to and attributes relevance to other stimuli, such as crushed vegetation, disturbed ground scent, or environmental and handler cues the less he will have attended to and learned about the subjects scent. Over time this problem can be compounded. Conversely if the hound receives preliminary training on the target scent, before being trained with multiple components to the task, then the amount learned with regard to the target scent will be greater. Given the multiple sources of cues and sensory information in a rich environment we can expect that the least ambiguous sources of information will have the greatest influence. If the subjects scent is more valid it will increase the willingness to attend to it and devalue the associative strength of competing stimuli. Competing stimuli includes input from all the senses, the internal mechanisms eliciting and governing behavior, environmental cues, or cues from the handler (Celenza, Kenny, Turkewitz, 1984; Stein, Meredith, 1993; Turkewitz, Kenny, 1982). What the hound senses, thinks, and does provides the aliment for future behavior. With continuous experiences of perceiving and acting expectations and representations are established which affects perspectives from other experiences. The sequence of experiences affects how future experiences are perceived.

Generally it is easier to add to experiences and knowledge than to modify convictions that are geared toward incorrect stimulus selection. The circumstance of the more salient stimulus selection is affected by the difficulty or the ambiguity of the discrimination task. Their ability is not necessarily a consequence of olfactory sensitivity. Animals do not necessarily attend to only one stimulus at a time and learning does not proceed independently of other available stimuli. However the associative strength of a stimulus and the behavior in response to that stimulus does affect the way learning proceeds. If the hound’s stimulus selection is flawed then so will be his development. New forms of behavior are conceived from his existing perceptions, and as associations strengthen so does the certainty of the associations. A competitive advantage can be gained by adding to prior percepts rather than spending time trying to change or adjust less effective ones. If you are trailing a person, by the way of your hound’s olfactory ability to discriminate the subjects scent from everything else, then the primary focus and the most relevant stimulus should be the subjects scent. Bad habits are harder to correct as they gain more and more foundation.

The hounds initial training and reinforcements should involve a minimum of opportunities for misunderstandings and distractions. His initial discriminative learning can be critically affected by the temporal and spatial separation between the various discriminative stimuli, the response to the target stimuli, and reinforcements. The things that make things easier to understand and remember assist learning. For example: the hound’s initial training might simply be to offer him a scent article. When he smells the article say smell and then immediately praise and reward the hound when he does so. His attention to the stimulus and the command paired with reinforcement will increase the value of the stimulus. This is usually established through highly positive praise and rewards rather than a neutral or non-reinforced response to the behavior. Non-reinforcement can detrimentally teach the hound more about the irrelevance of the scent. Stimuli that are designated as rewarding, irrelevant, or aversive affect the probability and degree of a response to that stimulus in a positive or negative manor. The important variable is reinforcement. Reinforcement orients observations to the regularities in the task that are related to the things of value. With continued practice the process can be reversed so the command smell comes before the act of smelling. Over a series of these closely paired exercises the hound will acquire additional associative strength between the command and the scent article. The next step would be to offer the article to the hound, give the command to smell, but not praise him for doing so. Instead have the subject in the vicinity. The hound should be momentarily confused about why he did not receive his expected praise after smelling the article, once the relations between the stimulus and events have been established. Attention is activated when something violates current expectations1. Because of the violation the hound will be in a prime position to make the correlation between the scent article and the subject. Learning occurs best when the hound makes the correlation himself. When the hound is observed making the correlation between the article and the subject praise should be offered to confirm the correct response and further strengthen the correlation.

The novice does not easily shift attention away from the highly salient elements of a stimulus situation to less salient but potentially more relevant stimuli. Because the presence of more salient scents or stimuli can interfere with learning about more elusive scenting tasks, if the hound lacks an established understanding of what is required of him, his initial training should not be on grass, soft surfaces, in highly contaminated areas, or ever on hot dogs distributed along the trail. Hounds initially trained with multiple scent components to the trail may learn less about the subjects scent and more about the track scent. The amount learned about the subjects scent during the hound’s preparatory training can greatly overshadow the potential of greater significance being associated with the more salient track scents, when the more salient components are later added to his training. The difference in the choice of discriminative alternatives may be a consequence of the different initial probabilities of attending to the relevant stimuli. The hound’s background can predispose him to a preferential choice of alternatives. If the subjects scent is more valid than the track scent than the subjects scent, over time, will develop more associative strength and consequently reduce the possible associative strength of the track scent. However, if the subjects scent lacks validity, to what the hound understands the task requirements to be, than the more salient track scent may gain more associative strength and validity than the subjects scent. And if the more salient track scent has been established first, before the subject’s scent as being the most valid, then the associative strength to the track scent may greatly overshadow efforts to teach the hound to the contrary (Kamin, 1969; Mackintosh, 1971, 1975). Overshadowing, in this case, hinges on the more salient component being authenticated as the guide to the task before the subjects scent, thus leading to a lack of associative strength to the less salient component, except at the end of the trail when a reward is anticipated. When the hound is trained to track crushed vegetation or disturbed ground scent the recognition of the subjects scent signals no change in the outcome of his goal and therefore represents no particular reason to attend to the subjects scent. What the hound is most likely learning, when trained to track, is more about the irrelevance of the subjects scent and hence will come to habitually ignore it. Over time as a behavior becomes more and more ingrained it can significantly reduce the willingness to explore other possibilities.

The hound’s pre-training effects what he learns when other components are added. Concomitant stimuli are not necessarily an inevitable correlate of recognition. Once the relevance of the primary stimulus has been established, the hound’s comparison between the relevant and the irrelevant stimuli can be an important component for the authentication of his directed attention. In other words; rather than associating things as belonging together, due to the pre-training the hound will more likely discriminate between the various stimuli present in the working environment. Pre-training to establish unambiguous identity of the relevant stimuli equips the hound to perform accurately on more potentially difficult discrimination problems. If the hound was trained from the start on unclear problems he may never learn the primacy of the relevant stimuli. The associative strength of the primary stimulus effects the way his learning proceeds. What is not attended to does not contribute to learning, whereas the things that receive pervasive attention gains associative strength. Associative strength of a stimulus is not simply a matter of salience. The associative strength of a stimulus is affected by prior experience, reinforcements both from the environment and the trainer, the salience of the stimuli, and the relationship to other stimuli. Once the hound has learned the consistent predictive relationship between the scent article (if you are training man trailing), the command, the subject, and the rewards his training can proceed to include a multitude of variables.

As the performance of a skilled task improves attentional demands from the novelty of the task decline, allowing spare attentional resources that can be directed to a wider range of things that are simultaneously present in the environment. If the hound accurately understands the task requirements, as his training progresses his experiences will alert him to a number of novelties to which he can validate causal inference during later training exercises. From prior learning experiences the hound is able to observe re-occurrences and consistencies, and form expectations. Different training locations and subjects, as well as honest thoughtful attention to the hounds development and understanding, will contribute to clarify what is relevant and irrelevant. The changing stimuli and associations formed from different contexts are crucial elements in the morphological process of perceptual organization, learning, and memory. Memory is context sensitive in that the stimulus situation evokes the memory. The representations or connections made from different contexts are key components to complex learning, perception, and the ability to respond timely and accurately. Each appraisal reveals correlations that may guide future behavior. The hound is constrained by what is not available to him, by features of both developmental context and his own characteristics. Perception is the process from which experience stands out.

The transitions from novel to organizationally based behavior and understanding are not necessarily age dependent or characteristic of a stage of development. The transitions are specific to the experiential history of the individual involved. There is no single learning or training process. It is important to understand that the experiential history of one individual can be outwardly similar to another’s but the level of desire to achieve a goal, and the amount or direction of attention paid to the experiences, combined with how they are comprehended, can result in vastly different outcomes. It is of equal importance to keep in mind that biological systems are dynamic; they continually transform. Understanding can vary with each individual, and heath, fitness, or coordination, as well as brain function, can strengthen or atrophy (see (Merzenich, Allard, Jenkin, 1990). There is no such thing as a finished dog or decontextualized trailing ability. It is true that an animal that is equipped with a more developed sensory system is likely to learn more about the sensory input than an animal with less capacity; however his ability is the product of biology and learning experiences. Learning about things or the environment is not just a matter of sensory input. Learning is facilitated or inhibited by prior experiences and the perspectives gained from them. The meaning something holds or the way things are mentally represented determines their applicability. Ability and meaning is constructed during the execution of tasks. It is a continuous process. Genetics may provide the potential for what they might become but their functional competence is a product of developed multiple sensory links within the hound and between the hound and the outside world.

The greater the hound’s cohesive functional adaptations the greater will be their capacities to interact successfully with fluctuating environments. These capacities dynamically assemble, strengthen, and evolve through continuous training and experience.


1. Young people and animals, and to an extent the novice, initially attend to and learn more about the things that seemingly belong together. (Bahrick, 1992) Once matching has been established perception and learning is guided more by expectations, desires or goals, and prior knowledge. After the initial matching, the things that are out of place or are related to a goal attract more attention. (see also Kucharski, Spear, 1985).


Bahrick, L. E. (1992). Infants’ perceptual differentiation of amodal and modality specific audio-visual relations. Journal of Experimental Child Psychology, 53, 180-199.

Celenza, M. A., Kenny, P. A., and Turkewitz, G. (1984, November). The effects of premature eye-opening on olfactory discrimination in the infant rat. Paper presented at the annual meeting of the International Society for Developmental Psychobiology, Baltimore.

Doubell, T. P., and Stewart, M. G. (1993). Short-term changes in the numerical density of synapses in the intermediate and medial hyperstriatum ventrale following one-trial passive avoidance training in the chick. Journal of Neuroscience, 13, 2230-2236.

Freeman, W. J. (1975). Mass action in the nervous system. New York: Academic Press.

Freeman, W. J. (1987). Simulation of chaotic EEG patterns with a dynamic model of the olfactory system. Biological Cybernetics, 56, 139-150.

Freeman, W. J. (1991). The physiology of perception. Scientific American, 264, 78-85.

Gottlieb, G. (1991). Experiential canalization of behavioral development: Results. Developmental Psychology, 27, 35-39.

Greenough, W. T., Black, J. E., and Wallace, C. S. (1987). Experience and brain development. Child Development, 58, 539-559.

Kamin, L. J. (1969). Predictability, surprise, attention, and conditioning. In B. A. Cambell and R. M. Church (Eds.), Punishment and aversive behavior (pg. 267-296). New York: Appleton-Century-Crofts.

Kelso, J. A. S. (1995). Dynamic patterns: The self-organization of brain and behavior. Cambridge, MA: MIT Press.

Kucharski, D. and Spear, N. E. (1985). Potentiation and overshadowing in preweanling and adult rats. Journal of Experimental Psychology: Animal Behavior Processes, 11, 15-34.

Mackintosh, N. J. (1971). An analysis of overshadowing and blocking. Quarterly Journal of Experimental Psychology. 23, 118-125.

Mackintosh, N. J. (1975). A theory of attention. Psychological Review, 82, 276-298.

McCabe, B. J., and Horn, G. (1991). Synaptic transmission and recognition memory: Time course of changes in N-methyl-D-aspartate receptors after imprinting. Behavioral Neuroscience, 105, 289-294.

Merzenich, M. M., Allard, T. T. and Jenkin, W. M. (1990). Neural ontogeny of higher brain function: Implications of some recent neurophysiological findings. In O. Franzn and P. Westman (Eds.), Information processing in the somatosensory system (pp. 239-311). London: Macmillan.

Rosenzweig, M. R., Bennett, E. L., and Diamond, M. C. (1972). Brain changes in response to experience. Scientific American, 226, (2), 22-29.

Stein, B. E., & Meredith, M. A. (1993). The merging of the senses. Cambridge, MA, MIT Press.

Sur, M., Pallas, S. L. and Roe, A. W. (1990). Crossmodal plasticity in cortical development: Differentiation and specification of sensory cortex. Trends in Neuroscience, 13, 227-233.

Thelen, E., and Smith, L. B. (1994). A dynamic systems approach to the development of cognition and action. Cambridge, MA, and London England: MIT Press.

Turkewitz, G., and Kenny, P. A. (1982). Limitations on input as a basis for neural organization and perceptual development: A preliminary theoretical statement. Developmental Psychobiology, 15, 357-368.

Williams, M., Johnston, J. M., Cicoria, M., Paletz, E., Waggoner, L. P., Edge, C. C. and Hollowell S. F. (1998). Canine detection odor signatures for explosives. SPIE, 3575, 291-301.


This paper is in its infancy and is subject to change.

November 2002 – Rev. 12/14/2003 6:10 PM