Scott of Tre'Good, Task Support Service Dog
OFA Good/Normal, MDR-1 N/N, DM N/N
(Coefficient of Inbreeding 0.0)
Scott of Tre'Good will be the sire of our 2018 winter litter. Scott is one of several exceptionally intelligent service dogs we have produced. Scott's full brother Haven is also expressing extraordinary talent and is an up and coming stud for 2019
Scott is a dream dog. He is safe in all public encounters, off the charts intelligent and very willing to work. Mike says living with Scott is like living with a human. His receptive language abilities are so developed that Mike can talk to Scott in sentances and he understands and helps as needed. Scott has all the qualities we strive for in our breeding program and we are excited to meet the pups this amazing boy produces. See below to read the preliminary results of a study being conducted on Scott by a Linguist.
We also use Scott as a mentor in our public access training program. Scott's confidence and focus on Mike makes him an exceptional mentor who easily transfers his skills to the young, impressionable pups.
Scott will be available for outside breedings with approved females.
Click on the pictures below to view more pictures of Scott
The video below was taken when Scott was 8 months old.
Below is an initial general summary of observations of Scott. This page will be updated when the study of Scott's liguistic ability is complete.
Scott’s level of obedience and facility with new commands easily place him within the highest of Stanley Coren’s intelligence rankings.
Scott does not know as many nominal categories as some of the other dogs studied in the linguistic literature, because he has been trained as a service dog, not an experimentation dog. But his facility with new object names and his nominal category knowledge are on par with such dogs (Rico, Chaser, Bailey, etc.) and include remarkable features not previously described in the experimental linguistic literature (apparent subcategory knowledge and possibly also ability to recognise category members based on shape rather than smell).
Scott learns new verbal commands with no training at all when they are given to him after he intuitively provides needed assistance in his capacity as a service dog.
Scott is able to recognise interrupted and uninterrupted versions of the same phrase as instances of the same command category, e.g. “Can you pick that up?” vs. “Pick up the ball.” (Note that “Up,” outside of this phrase, means “jump.”)
As a service dog Scott knows a great variety of verbal categories as commands toward relatively specific embodied actions. His extraordinary intelligence shows in two areas:
relatively diffuse verbal categories
Scott has an extraordinarily diffuse representation for the commands “open” and “close”. He is able to correctly perform these actions on a variety of targets, such as drawers and doors, which require very different embodied actions to open and close, and which have very different physical structures. The only natural affordance, it appears to me, which these objects have in common is that they contain sealed-off areas to which joint attention and activity can be directed only after the contained is unsealed. I did not observe Scott opening or closing an unfamiliar object, but I did observe him opening and closing a very difficult object – a sticky and weak drawer in a lightweight and loose dresser. He tried the tops and sides of the drawer with his claws, and failed to open it, before attempting to use his teeth to open it by the handle, upon which he succeeded. During this trial he refused to retrieve an object from the top drawer, apparently because he knew he would topple the dresser if he put his weight on it. After Mike moved the object to the bottom drawer, Scott retrieved it. Mike also pointed out that Scott closes drawers using the tops of his claws and thus doesn’t scratch them. It is unclear to me whether Scott expands the extension of “open” and “close” when he learns to open and close unfamiliar objects, or instead has a more human-syntax-ready knowledge of these categories and associates the different actions with the objects rather than the commands “open” and “close”.
acquisition of new action names
As Mike puts it, Scott figures out routines for helping, and Mike names them. At one point when Mike had a hard time getting up, he looked to Scott and issued an unfamiliar command, “Help.” Scott intuitively approached in such a way that his left side pressed against Mike’s right leg and his chest pressed against Mike’s left leg. Mike then asked, “Ready?” and Scott intuitively prepared to bear Mike’s weight on his shoulders. I observed this sequence once during each of my observations. Mike assures me that he never “taught” Scott either signal. Generally, Mike tells me, the learning sequence is such that Scott offers help when Mike encounters difficulty or appears to be doing something risky, and if Mike names Scott’s help offer using a command verb Scott learns the new word immediately.
Scott knows a number of nominal category names, including among my observations “ball”, “shoe”, “sneaker”, “boot”, “bully stick”, “toy”, “frisbee”, “food”, “treats”, “bowl” and “sandal”. He shows knowledge of these names primarily by attending to and retrieving named objects (out of an array of other objects) on command.
Scott knows the names “Mommy” for Autumn and “Daddy” for Mike. He shows knowledge of these names by bringing objects to or going to the appropriate person on command.
Scott’s nominal category knowledge in the case of “shoe” has a structure that has not been described in previous “dog linguistics” literature.
First, Scott uses nominal categories exclusively in order to display the attention and retrieval behaviors described above.
Second, Scott has apparent subcategories (“boot”, “sandal”, “sneaker”) and uses higher categories inclusively of those subcategories, correctly retrieving sneakers, boots and sandals when told to retrieve a “shoe”, despite the fact that he knows them as “sneaker”, “boot” and “sandal” too. This knowledge may well be parallel rather than hierarchical, but either way it has not been previously described to my knowledge.
Third, Scott uses nominal subcategories exclusively, correctly retrieving boots on command from a test set including other shoes.
I did not test for human-like implicatures arising from referring to “shoes” using the same test set. This is because Scott is ordinarily rewarded for retrieving boots when told to attend to “shoes”, and to change this reward structure could harm his sense of security and knowledge.
Twice during my second observation, Scott retrieved a toddler-sized Dutch klomp (wooden shoe) when told to get a shoe. These klompen had never been worn by human feet and consisted of painted wood, having little in common with other shoes Scott is in the habit of retrieving. It would be remarkable indeed to be able to confirm that Scott knows klompen are shoes, but further confirmation is required at present
Scott is able to identify the correct referent of phrases with the word “other” in them, either followed by a familiar noun or by the proform “one”:
for paired objects (e.g. “Get my other shoe.” / “Now the other one.”)
for comembers of a larger category (e.g. “Get the other ball.” / “Ok, bring me the other one.”)
acquisition of new object names
With gestural manipulation of joint attention, Scott is able to learn a new object name with a single repetition.
Without gestural manipulation of joint attention, Scott is able to use process of elimination to associate a novel name with a novel object.
Mike tells me this is a typical way of teaching Scott new names. I observed it during our first observation when Scott correctly retrieved the “orange ball” and not the frisbee or red ball.
Scott is able to comprehend such a new object name later without reintroduction.
ex.: During my second observation, Mike placed his boots and a favored black ball on the floor, then told Scott, “Get me my boots,” pointing towards the three test objects. Scott retrieved one boot, and Mike took it and then put his hands under his legs and looked down, adding, “Get the other one.” Scott retrieved the second boot. That same morning was the first time Scott had been given any command with the word “boot” in it.
ex.: During my second observation, the four of us were standing around the kitchen bar. Mike asked Scott, “Where’s Mommy?” and Scott approached her then looked back at Mike. He then asked Scott, “Where’s Bryan?” and Scott turned to me and jumped up on me. My name had not been mentioned since the first observation, 5 days prior.
Scott is able to use human-language modifiers of known object names in order to disambiguate. He can regularly distinguish between “red ball” and “black ball”. I did not observe him distinguishing between “Mommy’s shoes” and “Daddy’s shoes,” but Mike tells me he can do this.
Most other modifiers, Scott appears to use in similar manner to new object names.
I did not test for the syntactic structure of Scott’s known modifier-noun phrases.
responsiveness beyond area of joint attention
After hearing an object name and the command “search”, if he is familiar enough with the object name, Scott is able to retrieve an object even without knowing where it is or being given gestural direction.
Bryan James Gordon, MA
Instructional, Linguistic and Technical Support
Title VII Umóⁿhoⁿ Language and Culture Center
Umóⁿhoⁿ Nation Public School