RETRIEVAL CUES AND CANINE LEARNING: PART ONE
Ever wonder why dogs work so well at practice but perform below-par or even disqualify at the shows? At practice, the dog appears to know the exercises and the handler has carefully reinforced and motivated the dog for performance; so why doesn’t the dog perform as well in the show as he does in practice? Performance depends on many factors including (but not limited to) learning, memory retrieval (remembering), and motivation. A dog may not perform in a situation because he is not motivated to do so, but he certainly can’t perform if he is unable to remember the information. Retrieval of learned information is necessary for performance, and there are factors that can aid or inhibit remembering.
Learning is defined as a relatively permanent change in behavior as a result of experience. Memory is an active mental system that encodes, stores and retrieves learned information so that performance can happen. Encoding is the process by which information is converted into mental representations that the brain can process. As information is learned, it is encoded in memory. Along with this main information, some incidental information regarding the physical context and emotional and biological state during this process is also encoded, and the encoded incidental information becomes retrieval cues that can aid or inhibit remembering. There are two types of retrieval cues which trigger the memory’s retrieval of this learned information:
Context-dependent cues contain information regarding the physical details of the remembered event or experience, like location, ring set up, grass, crowds, specific people such as the handler or judge, noise etc.
State-dependent cues contain information about feeling, mood, attitude (happy, playful, depressed, anxious, afraid) and biological states (gut reaction, stomach in knots, butterflies, goose bumps, exhausted, energetic).
That means that the learning desired for the exercise is encoded along with cues regarding the physical context and emotional state from the circumstance within which the dog learns. These encoded cues could be physical details of the environment and moods and attitudes experienced and responded to. State and contextual cues trigger remembering of the main learned behavior and have a significant impact on training and performance.
Cues present during training may be very different from those present at a show. Training context is less cluttered and distracting, and is often quieter than the noisy barking, clanging and announcing at the shows. Handlers’ relaxed and confident training state gives way to nervous uncertainty at the shows. Even our training and reinforcement style, encoded as retrieval cues, can be different from those presented at the show.
Learning is a dynamic process. Throughout our conscious waking time, we are learning. (OK, well maybe not every second.) Our dogs are learning all the time too, and incidental retrieval cues are being encoded along with this learning. Some factors that can affect new learning are; prior learning, number of exposures to, and significance of events. Remembering is an interactive process involving the physical context, environment, handler, dog, behaviors, etc., and the mental and biological states that change and evolve as the situation updates.
The interplay between the participants during teaching and learning is intensive and reciprocal. Teacher/trainer actions and moods, coupled with the dog's reactions to them may be encoded as cues for the dog. The same is true for the teacher/trainer, who responds to and learns from the dog. Emotion, attitude and reinforcement during training is encoded along with learning and can be reflected in performance.
Information retrieval functions best when the learning state and performance state are the same. Minimize the difference between training and show by simulating the show situation in training and the training situation in matches/shows. Include the ring setup, noise, crowds, mood, training style and reinforcement/praise. This does not mean that teaching and learning must be accomplished within the rigid requirements permitted at a show. It means that learning is progressive, each step leading closer to the performance state, and minimizing the difference between training and performance.
Some cues trigger memories of the desired learning while others trigger incompatible or competing memories which inhibit or prevent retrieval of the desired information. Similar training and performance situations present similar retrieval cues which aid remembering; but when situations are dissimilar, contextual and state dependent cues may not be present to aid retrieval. Instead, the present cues may not be linked to remembering the desired information.
Example: In practice a dog has only learned go-outs only to poles, then he finds he must perform in a ring with baby gates. The baby gate contextual cues are very different from the cues for poles, and would not aid retrieval of the correct go-out information because that information is linked to the pole contextual cues and not to the baby gate cues.
Dogs may perform in a wide variety of circumstances; outside at parks or racetracks, indoors at convention centers or livestock pavilions, and each show has its own unique site picture, smells and ambiance. Helping the dog to learn in as many different situations as possible will maximize the effects of retrieval cues. This method of learning is encoded with many different sets of retrieval cues from many situations and states, allowing a broader based knowledge with more varied access to the information. It is not dependent on one set of cues for retrieval.
Retrieval cues may also inhibit performance by accessing competing or incompatible memories.
Example: A puppy's initial training consists only of the "sit" command rewarded by food. The learned "sit" information is encoded along with the contextual retrieval cue of food presentation. Soon, presentation of food is enough to retrieve the sit information and the pup sits regardless of what verbal command is given. Later, this same pup starts conformation training. The objectives are to learn to stack and bait, which require that the pup stand still (stack) and focus on a food reward (bait). The contextual cue of food presentation encoded earlier will trigger retrieval of the "sit "information and the dog may/will sit. Interference from contextual cues may explain why many conformation people do not teach their dogs to sit.
Interference is possible with cues encoded from corrections (punishment) as well.
Example: Envision teaching the stand-stay of the signal exercise using corrections whenever the dog moves. The dog learns that moving results in a correction and "don’t move" information is encoded with the contextual and state dependent cues present. So far, so good...but once the dog masters the stand, he needs to learn to drop from it, and because the situation is so similar to the learning situation for the stand, the context and state cues trigger the ‘'don’t move" learning, so the dog does not drop! Retrieval cues have interfered and retrieved competing learning.
Cues encoded during the learning process can cause interference.
Example: Sure, we all know that the command word “down” is a retrieval cue triggering remembering of “down” information, but many other cues are processed by the dog during learning along with the command word; a pull downward on the collar, sliding the legs forward from a sit, pushing on the shoulders, position and presence of the food lure, location, mood, etc. These cues, as well as the verbal command, may all be encoded with the desired information. Techniques and manipulations performed in order to elicit the desired behavior may also be cues for retrieval of that information. It is through a series of steps that the dog learns that the word is the specific cue (among many) that requires response. Sometimes dogs get “stuck” responding to one of the teaching cues (like the food lure or physical manipulation) and won’t respond to the verbal command. Repetition and time will sometimes make the connection. Withholding the teaching cue (lure and lure action) and repeating the command while waiting for the desired response before rewarding will often work. If not, changing technique may add and change cues providing additional access to the main learning.
What’s the role of rewards? The information we want has already been remembered, so what significance could the reward play in information retrieval? Rewards, or lack of rewards, may be linked to cues affecting the dog's confidence level.
Example: At practice, dogs are rewarded for performing correctly and are not rewarded for incorrect or non-performance. If reward is usually given when the dog is right, the reward may become a cue linked to the dog’s feelings of being correct - the “I’m right” feelings. Since incorrect performance is not rewarded, absence of a reward may be linked to the “I’m wrong” state. If the rewards used in training (and those expected by the dog) that trigger the “I’m right” state are rewards not permitted in the obedience ring (food, toys), problems can occur. The show ring permits rewards of praise and petting and requires that the dog be under control at all times (how boring!). So, handlers reward with praise (frequently inhibited), some petting, or do nothing at all…no reward. The cues for the “I’m right” state (toy, food, exuberant handler) are not presented; instead, unconvincing praise or silence provides cues that may trigger the “I’m Wrong” state. Different dogs react differently to an “I’m wrong” state; some might become unsure, less confident. Performance may slow down, diminish or even stop. Others may respond to this state by becoming anxious or frantic, winding up and trying so hard they do the exercises nearly out of control.
handler mood, apparent in most show situations, may also trigger access to
learning not desired in those situations. Tense, serious moods may trigger
learning linked to anxiety and worry in the dog; perhaps because of a previous
unpleasant exposure to the same mood paired with an uncharacteristically harsh
correction, an angry training session, or quick rebuff around the house, and
resulting in the dog's hasty retreat. Repeated exposure to cues eliciting unsure
or anxious response in the show environment can result in a dog that “shuts
down” at or does not like shows.
The handler’s actions and moods (states) when teaching the main learning may also be triggers for the dog’s remembering. If learning was encoded along with negative feelings or physical discomfort (his and/or ours), similar situations/commands/exercises will trigger those same feelings and biological responses. Dogs learn all the time, not just when we are training them, and encode information and cues everywhere, not just in practice. Pleasant or unpleasant events and their resultant feelings and states are learned and encoded. Similar contexts, exercises or performance situations can trigger access to pleasant or unpleasant memories that can enhance or stifle a show career. Negative experiences at shows may be encoded with show context and state-dependent cues that aid access to those unpleasant memories at shows.
Manipulation of retrieval cues gives broad access to information. Some dogs are required to perform in a wide variety of circumstances, outside at parks, indoors at convention centers or fairground livestock pavilions, each situation has it’s own unique site picture, smells and ambiance. Maximize the effects of retrieval cues by helping the dog learn in as many different situations as possible. Learning encoded in different scenarios will have many sets of retrieval cues from many situations and states. This is broader-based knowledge with more varied access, and is not dependent on one set of retrieval cues.
Learning is encoded along with cues pertaining to the contextual, emotional or mental state and biological feelings. Understanding how retrieval cues affect remembering permits training programs to be carefully managed for best results.
Examples illustrate some of what happens during the learning/performance process. Descriptions are simplified for clarity and do not condone or criticize any training methods or models.
Copyright Debby Boehm. DO NOT REPRINT WITHOUT PERMISSION.