The Norbert experiment, part 5: a session with the new adjustments goes well … even though this cat is being difficult!

This cat is closer than the baseline cat. It’s not always visible, but Game knows it is nearby, and is tempted to go back and figure out exactly where it is. I think today’s video is fairly self-explanatory since I’m chatting about what’s going on throughout. If it isn’t – ask away in the comments.

I’m keeping this video short – cut out a long part of our tug play and the end, but: I could easily guide Game into the bedroom (which is a small dark room, and that’s the reason we don’t usually hang out there), close the door, and relax with her on the bed for a bit, heading back out into the other room when the cat was gone.

In part 6, I’m going to share how I will try and keep something resembling data on these sessions. Not as clear cut as I’d like, but it’ll have to do, and I think it will work for our purposes. Stay tuned!

The Norbert Experiment, part 4: adjustments

So much to learn from yesterday’s session!

The reason I ran out of both high-value treats AND kibble is that I did not expect to go more than 2 or, at the most, 3 sessions. But such is life! I learned:

  1. As long as the toy is out, Game will keep going.
  2. Game is indeed able to choose the toy over the cat after running down the staircase several times (her motivational state is likely different at that point due to being hot and tired because neither of us is used to this climate).
  3. Game can take kibble within this staircase marker game – at least starting at the 4th round.
  4. Once in the bedroom after the 7th scatter (no sight contact to cats or toys), Game is able to relax right away.

Based on this, I will adjust in the following way:

  • Remove the toy during the scatter (leaving it on the floor until then will give Game the option to grab it if she needs to earlier on).
  • Only go down the staircase once.
  • Move swiftly to the bedroom after the scatter and take a break – no matter whether there are cats or not.

Making it measurable

I’m going to do a session like this, and then decide how to best measure our progress. I also realize this is not necessarily going to be easy to measure because there are different cats, and the distance at which they appear often differs, too. Sometimes they move, and sometimes they are stationary. Sometimes there is only one at a time – and sometimes there is more than one. Sometimes, they stare at us (we are in a display window at this AirBnB – for people and cats). This means the sessions are not directly comparable, which is a shame. We don’t live in a lab!

If anyone readong along wants to throw their suggestions of how they’d measure this at me, go right ahead! I’m writing this up more slowly than I’m training, so by the time I read our suggestion, I will likely already have implemented whatever I came up with myself. But if you want to think along and decide what you would do in my situation – go right ahead and have fun with this in the comments, and I’ll be sure to get back to you there!

Since I have no video for you today – take one of the cats instead:

The Norbert Experiment, part 3: a long session filled with information!

I’ve dug the harness out of my luggage – still looking for the longer line (when you live a mostly off-leash life, you tend to lose track of your lines). We’ve been on a long walk, it’s hot (significantly warmer than at the previous place we were), and Game and I are both tired. Lazy play is just right for both of us today!

Here’s Game’s marker/arousal staircase image, with time stamps for each of the steps below.

Video A

Round I

Step 1: Consider the lobster cat

00:00-00:10 Considering the cat (longer than ideal but I wanted to show you all the cat)

Step 2: Tug

00:10 Tug marker. Game responds well!

00:30 I realize I had closed the glass door and can unclip Game from the tether.

01:05 A quick look at the cat, and then Game disengages and keeps playing! Yay!

Step 3: Chase high value food

01:37 My first tossed treat cue. Game is slow to let go of the toy here! These are the things I pay attention to: does she respond to marker cues at baseline speed or below? This response is below. I may have caused it by my teasing tiny tugs on her toy right before the marker though. It is not clear whether the latency is cat related.

01:45 I remember I was going to leave the toy out, and see whether Game will gravitate towards it if she needs to sink her teeth into something after the first run down the staircase.

01:50 I was not planning on tossing the treat at an angle that would let Game see the cat easily right after eating – she does really well though, and does not get stuck.

02:13 Team work!

Step 4: Eat high value food from hand

02:16 First click for eye contact!

02:25 That look may have been at the cat (who is still under the white table) or the person walking past. I can’t tell.

02:29 This look is clearly at the cat – the person has passed already. This is okay: looking at things in the environment/pointing them out by looking is just as clickable as looking at me. This is a both/and, not an either/or paradigm.

02:34/35 Another one for looking at the cat! You just look where you need to look, Game.

02:37 I shift a little to make the different directions of looking more salient. Now, looking at me is clearly looking away from the cat, and looking at the cat is clearly looking away from me.

Step 5: Scatter

02:46/47 First scatter cue, marking eye contact.

Round II

Another round! Game is looking at the cat again. Calmly so – but she’s looking. For now, I will take this as a cue to restart the staircase.

Step 2

03:32 Tug cue. Game is looking at me (she knows what’s coming), and I can mark that. If she were to continue looking at the cat, that’s what I’d mark with “Tug!”

04:27 By moving away and allowing Game to bring the tug toy back to me, I’m giving her the opportunity to restart the game. This is how I keep things cooperative.

Step 3

04:40/41 That response to my “Get it” marker was perfect: that’s Game’s baseline speed, and what I want to see! As soon as I said, “Get it,” she let go of the toy.

04:45 Again, tossing into the cat corner is not what I had in mind when coming up with my training plan.

04:47 … but Game handles it well! So well, in fact, that I might want to add a sprinkle of the Give Me A Break CU game to our cat sessions!


Sidenote: this is what Give me a break looks like: the treat can eventually be put down close to a stimulus, and the dog will dismiss, like Game dismisses the police person in the video below:


Step 4

04:56 Click for eye contact.

04:58 I’m moving to give Game more of a choice in terms of whether she wants to look at the cat or at me – now we are obviously in different directions.

Step 5

05:11 Scatter cue for eye contact.

Round III

05:37 The cat is still there, and Game is watching again. She is not in predator mode (which may be because of the scatter … or because she isn’t used to the heat. Being hot (or cold) influences motivational states.

We start over with step 2 after video A ends.


Video B

Round IV

Step 2

00:00 The last part of toy play. Game went back to cat-watching after the previous scatter, and I started over at step 5.

00:08/09 You can see she’s tired. She isn’t super fast and intense, and holding the toy gently rather than hard. But keep going she wants, and I want to keep experimenting – so we keep going.

Step 3

00:26/27 Realized the treats from my pocket were gone; had to get them from the counter. At this point, we are using kibble – I have gone through all the high-value treats I cut up already.

00:37 Game doesn’t mind chasing kibble – this is good!

Step 4

00:58 Waiting for Game to offer eye contact …

Step 5

01:11 Scatter cue for eye contact. Kibble again. Game doesn’t mind.

Step 6?

01:39 She starts circling here – she considers laying down for the first time!

Round V

Step 2

01:43 Then she circles past the tug toy. This is a training toy, not a toy I usually leave out for her to disembowel. She can’t resist it, and asks for another round.

This is good information for me: she did not look at the cat, and then choose the toy. She was going to do step 6, then saw the toy and changed her mind. I’m going to need to adapt this approach (leaving the toy out) since this is not what I’m aiming for.

01:50 I mean it’s a good decision to bring me the toy rather than get sucked into cat watching. But watching this video back, I can see that the decision she made was not “do I stare at the cat or get the toy,” but “do I lie down or get the toy.”

We continue down the marker cue staircase again after video B ends.


In round 6, Game looked at the cat, and then channeled her cat thoughts into the toy unprompted – that is awesome and exactly what I was going for! Good girl! I’m not showing you video of this because by that time, the camera had fallen over.

There is a 7th round. Rounds 4, 5, 6 and 7 were all played with kibble rather than high-value treats. By round 7, I run out of kibble as well (I only got her portion for the day from the car, and have no refill at hand). So after the round 7 scatter, I encourage Game to follow me into the bedroom and close the door (no sightline to the cats). She is able to relax right away.

Lots to learn from this long session! Tomorrow, I’ll share the adjustments I’ll make based on what I’ve seen in this session. There is lots of room for our cat experiment to grow!

The Norbert Experiment, Part 2: cats happen faster than expected

This is a short one, just showing you how messy real life is! I filmed the baseline pretty much right after getting to this AirBnB and seeing the cats. Then I tethered Game to the couch and opened the glass door so there was only the screen door because it was HOT inside! I was still getting set up here myself, but managed to prepare some of what I would need for cat sightings (chop up high value treats) by the time this cat – the next one after the baseline – walked past. Both the longer leash I’d rather use for Game’s tether and the harness I’d prefer to tether her on (as opposed to a collar) are still doG knows where in my luggage or in the car. They’d have been up on top had I known how cat-y this AirBnB was going to be.

So life happens, and we make the best of it!

Below, as I was still digging through my luggage, a cat walks past. I record the cat (rather than attending to Game right away), and then I quickly run down the marker cue staircase.

The good news: it works well and Game is able to respond to the tug cue right away.

The not so great news: by the time I took this video, I had not had time to think through the details. For example the tug play you see in this clip is not what I later decided I wanted, and then described in part 1 (cooperative tug): here, I am reaching for the toy way more often then I’d like. The leash is too short to play on well, and I end up tossing a treat towards (rather than parallel to) where the cat used to be – outside of Game’s leash radius. And for most of this, Game’s not even in the frame of the camera. Oh well!

But after that scatter, the cat was gone, and Game was indeed ready to move on with her life. So while by no means perfect, the basic structure is working – and we’ll be better prepared next time! (With the camera set up in a good spot and ready to record, high value treats and toy ready to go, a longer leash and a harness!)

In the video below, I pick up the toy when moving to “Get it” treats. I might experiment leaving it on the ground next time so Game herself can restart the game if she needs to channel some more cat feelings into it.

The Norbert Experiment, part 1: Game’s cat baseline

We got to a new AirBnB the other day. Turns out there’s a lot of cats in the shared yard space the apartment opens out into! It’s also warm here, so for the most part I’m only closing the screen door, but not the actual glass door. Screen dors are not Malinois-proof barriers, meaning I need to tether my dog to keep the risk of cats (or screen doors) being harmed as low as it can be.

I’ll be here just long enough to turn this into a fun little training project. I’m finding this project particularly intriguing right now because I’m also working with someone on household-cat-acceptance using a different (tried and true) approach – more in that below. The approach I’ll be using with Game in The Norbert Experiment (1) is a bit more experimental, and it’ll be fun to compare the two.

Baseline response to tongue click, treat toss cue, and “leave it”

Today, I’ll show you Game’s baseline response to cats outside the apartment (glass door closed here, not just screen door). I know she wouldn’t be able to take the treat from my hand (this is kibble; I can usually work with it on almost anything). I’m only clicking and offering it to her to demo to you all that she can’t take it, not because I expect her to take it.

I was not sure if she would be able to chase kibble. Chasing food is higher value and higher arousal for Game than the same kibble from my hand, and she is able to do so in many situations, even when she can’t take treats from my hand. Game says, nope, can’t do.

I am surprised that she is able to respond to my “Leave it” cue – twice, no less – in this clip, but maybe I shouldn’t be: I have reinforced “leave it” as well as recalls with permission to chase cats, and that is the functional reinforcer Game is after in this situation.

From this clip, I learn two things:

  1. I need higher value treats for this project.
  2. I’ll want to consider adding toy play to my reinforcement/arousal shifting approach. I suspect (but will have to ask Game to know for sure) that tugging is equally high value as considering out-of-reach cats.

Speaking of reinforcement value and dogs who like to move their body …

Game could not chase treats here, but Keeshond Via below sure can. I love this clip because it shows that for Via (who just saw a deer), chasing treats is possible even when taking treats from Allison’s hand (“Yes”) is not. I love marker cues! I also believe we are severely underutilizng them in the dog training world. (Keep your eyes out for anything Karen Deeds has to say about this topic!)

Thank you, Allison, for allowing me to share your clip!

Chasing as a reinforcer (for coming back, leave it-s etc.)

With the cats in the alleys of Guanajuato, I used chasing as a reinforcer, just like I do with birds or squirrels who can easily get to safety. Guanajuato’s alley cats are dog-savvy and know that they just need to jump up a wall or roof and can give Game the finger from up there. I did not feel like I was adding substantial stress to their lives – just a single jump, which is something they are used to doing in their environment. I know this is an ethically foggy area. Personally, I’m okay with chasing as a reinforcer as long as it does not (and this assessment will be subjective, too) unduely stress the animal(s) being chased.

Here’s Game chasing birds (I don’t have a video of her and an alley cat). You can see how when I start the video, she is not mindlessly going after the cattle egrets. Quite the contrary: she waits for me to ask something of her so she can earn egret chasing. And as the video continues, she gravitates towards focusing on me rather than the birds. It is SO powerful to harness your dog’s greatest distraction, and turn it into a reinforcer! It removes the conflict of either/or (either I do what my human wants, or I do what I want) and replaces it with a both/and paradigm.

LAT on a mat for cat acceptance

The team I mentoned above is currently doing LAT from a mat, with the ultimate goal being acceptance of the household cat: the dog is on a mat, and gets marked and fed in a specific way for either pointing out the cat to her human or offering eye contact to her human. We keep the sessions to one minute and track the looks at the cat versus looks at the owner, and have a certain threshold point where we reduce the distance between cat and dog by one carpet square. This method is tried and true, and should also work for the goal of the dog learning that she will never chase that cat (anything CU means there will be no direct interaction with the stimulus). Once the dog is aware of this, it removes a lot of uncertainty from interactions. Uncertainty is stressful for many dogs, which is one of the reasons CU can be so powerful.

This is what this looks like with Heather’s cat Vignette and her Dutch Shepherd Saphira:


Sidenote to stress the beautiful training in the LAT clip above

Note that there are a several foundation behaviors that go into a training plan like the one Heather, Karl, Saphira and Vignette are implementing. If you don’t know what to look for, this may look effortless – but it is, in fact, amazingly complex and based on strong foundation behaviors, clear communication, and an excellent dog trainer (Heather) working with an excellent cat trainer (Heather’s husband Karl). Not only are they really good at what they are doing with their animals – they have also managed to build a habit for themselves: the habit of working on this together every evening they are home.

A training plan that requires time usually also requires us humans to develop a strong habit to keep at it. We can trick our own minds a bit here by clearly defining when that habit is going to happen, and turning an already established preceding event or behavior into our prompt for that new habit. Heather and Karl have been making great progress because they have committed to working on this at a particular time every day. The more we make something a habit for ourselves, the easier it feels to make time for it, and the more progress we are going to make.

Saphira’s foundation behaviors that need to be in place before a session like this is even possible:

+ A strong station-on-cot behavior

+ Understanding that the marker cue “ground” means a treat will materialize on the ground in front of her (in this case, on her cot bed, between her paws). The reason this is the marker cue we chose is that it resets Saphira: when taking a treat from between her paws, she is looking away from both Vignette and Heather, and can then make a new choice: does she want to look at Heather or does she want to look at Vignette? Both behaviors get reinforced equally.

+ The LAT game in easier contexts (knowing that pointing out a stimulus in the environment is a reinforceable behavior).

+ Knowing that eye contact with Heather is a reinforceable behavior.

Vignette’s foundation behaviors:

+ Relaxation in the presence of dogs.

+ Wearing a harness.

+ Walking on a leash and harness voluntarily. Karl isn’t pulling Vignette up to Saphira – they are just walking up together.

(The leash and harness are for safety, to make sure Vignette doesn’t run up to Saphira, just like Saphira is wearing a leash that Heather holds for safety. We don’t plan on needing them, but it’s good to have them – like seat belts.)

Shout out! Heather, Karl, Saphira, Philo (their second cat) and Vignette – I LOVE the work you’ve been doing, and the progress you’ve been making! You make a fantastic team!


Back to the Norbert Experiment!

Since I don’t mind sending Game to chase some of the time (I just need it under stimulus control), I don’t need (or even want) a purist CU approach here. Instead, I want to marker-cue her down every time she sees a cat outside of this particular apartment.

The idea is that eventually, seeing a cat through the screen or glass door will become a cue to ask me for a toy. This may not happen before I move out of here – I’ve got about two weeks. OR it just may. We’ll have to find out! I enjoy playing prediction games with myself, so I predict that I will see some kind of progress in these two weeks.

The image below is what I suspect Game’s hierarchy of arousal and reinforcement value looks like. She might proof me wrong, which is okay. I’ll have to ask her in order to find out if we’re on the same page about this! I suspect that considering cats is as arousing AND as valuable as tugging with me. The shift from step 1 to step 2 is therefore a horizontal one. It is not a shift in reinforcement value or arousal level, but a shift in attention (from cat to toy/me).

From step 2 onwards, I can then – theoretically – go down the arousal and reinforcement value staircase, shifting vertically down from one step to the next lower one until we are at step 6 and can move on with our lives, not thinking cat thoughts.

I may be wrong about the value of considering cats outside the apartment versus the value of tugging in their presence. That is okay – I am going to ask Game if this reinforcement hierarchy is indeed hers, and adapt based on her response.

I’ve used this method with different dogs in the past, but this is the first time (and I might misremember because human memory is not reliable) that I consciously include a potentially competitive toy game: tug. Tug can easily turn into a zero sum game, which would further increase arousal. I’ll try hard to keep it cooperative rather than competitive. I do not want Game to fight me for the toy. In order to keep it cooperative, I will make sure to keep letting go of the toy and allow Game to restart the game. I will also push back against her chest at least as much as I’m pulling on the toy, and I’ll work with her on the floor rather than standing up. This way, I hope to play Game’s calmest version of tug.

Should there be toy play around cats?

I’ve thought about whether tug is a good idea at all, because it will likely keep Game’s arousal at cat-level, and it is directly related to sinking her teeth into a toy – someting I would very much not want her to do with a cat.

Since this is my own dog who I like to experiment with (and know I can keep the cats safe from), I am going to go ahead with it and find out what happens. I believe, based on Game and my history of toy play, that this is going to increase her impulse control around cats – and that’s what I want. I don’t need relaxation right away, but I want cats to mean tug.

So at first, I will take her focus vertically from cats to a tug toy. Then, I’ll bring her arousal level down horizontally by switching from tugging to chasing high value treats, from chasing treats to treats from hand for offered focus (“Can you offer a behavior with a strong reinforcement history?”), and then to a scatter. Sniffing for treats is a relaxing behavior. Seeing whether Game can or can’t engage in it will make an excellent gauge of whether she is able to move on. Since most (not all) of the cats out there walk past rather than staying right outside staring in, by the time I’m all the way down to the scatter, the probability that Game will be able to move on (because there is no more cat) will be high, too. If not, I’ll do another round.

Keeping data

I’ll need a way to track my progress or lack thereof. My preliminary plan is to switch from high value treats to kibble every fifth time I go down this staircase. Will Game be able to do it with kibble or not? I’ll also keep recording video after the scatter to see what happens, and find a way of coding her body language to know how long it takes her to truly move on (maybe how long it takes her to lie down in a relaxed position and not stare catwards.)


(1) I’m calling this The Norbert Project because I just met Kayla’s cat Norbert, and sadly, I could only invite Norbert into my previous AirBnB when Game and I were out. It would be nice if the next time Norbert and I crossed paths, he could actually be inside at the same time as Game – even if only for a little bit. Also, meet Norbert, travel cat with Kayla of K9 Conservationists. A shout out to Norbert for inspiring the name of this training series, and for being his amazing van life travel self! Cats don’t get much cooler than that!

Body language in free-roaming dogs, or: meet the dog in front of you!

This is the full version of the description that goes with today’s Youtube video on the Free Ranging Dogs channel. If you’ve read the first part of the description already, pick back up under the heading “Dog #4”! If you haven’t – here’s the video description from the beginning:

Game is happy to be allowed to run off leash again (nothing to worry about – the surgery I mention in the video was minor and all is well, but she’s only been out on leash for the last 2 weeks).

This video shows how, in just 3 minutes, Game meets 5 different owned free-roamers. Just like pet dogs differ, so do the personalities, looks and behaviors of these dogs.

All 5 free-roamers in this video are owned dogs. That is to say, they live in the respective yards they come out of. Their gates are always open. This street is part of Game’s home range and part of the other dogs’ core area. Only dog #4, the Doberman/Lab (this is not a Doberman; I’m just picking look-alike breeds for you to distinguish them) is not inside their own yard – the person working on the car is probably their human, and the dog is out here with them.

Dog #1: the Husky

Game sees the Husky before I do. At 00:29, she greets them with a friendly wag and moves on. What you see at 00:29 is a behavior that lets me know there is a dog to her right.

00:36 The Husky comes out – hackles up at first, but Game has already moved on, so the Husky doesn’t care. Instead, they show curiosity/interest in me, and their hackles come down all the way.

Dog #2: the big black-and-white pup

00:54 This one looks pretty young to me – but I can’t say for sure; he may just look that way because of a recent hair cut. He, too, comes out of his territory. Unlike the Husky, the pup is interested in Game: friendly, waggy and playful.

01:02 Game responds to the friendly interest the pup is showing. She may be in the mood to run together. That’s because she’s been deprived of exercise for the last 2 weeks, and I’ve also mostly kept her away from from other dogs. When that happens, she tends to act more playfully until she’s back to baseline in terms of exercise and intraspecific social interaction. It usually takes her a few days to get back to baseline.

01:16 Game would have pooped here, but because the pup is still there and being playful/friendly, she forgets about pooping and reciprocates the playfulness.

01:19 Btw, the ear position you see here in Game – ears up and turned back – is what she’ll usually show when we’re out and about. This is not a sign of insecurity, submission or fear. (Game can do a whole bunch of things with her ears; this is just one of her many expressions.) She’ll usually have her ears up and back like this when she’s ahead of me. She watches what’s up ahead, and keeps an ear on me at the same time.

Here, she’s running towards me, and her ears are up and back to keep an ear on the pup who she’s allowing to chase her. Ears up and back are a sign of split attention in Game: eyes in one direction, ears in the other one.

01:24 … and running back the other way, in exactly the kind of speed that is right for the pup (who seems to have a hurt paw/leg and is not super fast). Game enjoys both being the chaser and being the chasee.

01:25 And yes, I say in this video that she’s been on limited activity for a long time. For me (and for Game, but really, mostly for me), 2 weeks are a fucking long time! Walking is my thing. And without a dog, it isn’t fun.

Dog #3: the Chihuahua

01:42 The Chihuahua has just come out of their yard, and wants to see what’s going on out here! Since the 5 dogs (the free-roamers) are all neighbors, the Chihuahua isn’t interested in the pup, but in Game.

01:45-01:48 The Chihuahua displays their interest by sniffing. They are confident and curious, and the fact that Game ignores them (“Too small; whatever; also I’m done playing”) likely raises the Chihuahua’s confidence to the bouncy, chasey level you see here.

The Chihuahua and Game aren’t playing – the Chihuahua is sniffing while chasing Game, who ignores them because she’s already on her way.

Game is aware of size differences and is much more likely to ignore a small dog than a large dog. The Chihuahua isn’t unfriendly, but not exactly friendly either.

01:49 Game may just have left the little one’s core area, making her less interesting and me (I am still in the core area) more interesting. The Chihuahua folds the ears back and wags at me in a friendly-submissive greeting gesture.

Dog #4: the Doberman/Lab

01:52 To your right, where the cars are parked, you’re about to see the Doberman/Lab. This dog is insecure and barky. They are in their core area (this is one of the neighborhood dogs here), but not in their own yard. They are likely out here with their human.

You’ll see the insecurity in the retreat and the continued barking:

02:01 Retreat.

02:04 Now that Game has passed, the dog is coming forward again: when one dog turns their back on another one, the other one will feel safer. Game just passed and ignored the Doberman/Lab.

02:06 … which is why the Doberman/Lab can now come forwards again and bark – this time at me.

02:12 The response to me is barky, but not fearful. There was only a fear response when Game was walking towards and past the parking lot – so this dog’s insecurity is dog-specific.

02:14 It’s hard to say whether the Doberman/Lab is in their territory or in their core area. In any case, the person at the car is probably their person.

It is entirely possible that the dog’s response to Game and I would be different if there was no other human present. Being with their human generally gives dogs greater confidence/perceived strength.

Dog #5: the second fluffy big one

02:19 This dog was probably alerted to Game’s presence by the barking of the Doberman/Lab. Like the Chihuahua, he wants to see what’s going on! He is not interested in me and runs out of his territory (yard) and right past me to check out Game. The barking you keep hearing in the background is still the Doberman/Lab, not dog #5.

02:28 Game is done socializing for this outing, which is why she isn’t giving dog #5 any attention. Dog #5 is just curious about her – no strong feelings in any direction. Having caught up with her, he sniffs where she sniffed, and later, he’ll pee on the corner of the wall.

This dog is confident, has no ill intentions, and is an adult. Among confident adults with good social skills, if dog A ignores dog B, dog B will also politely leave dog A alone. (There are exceptions. Sometimes two adult dogs – just like humans – dislike each other at first sight. But that would be an exception for socially confident good communicaters. Politeness is the rule: live and let live.)

02:39 You can see dog #5 pee and look around (for example at me) with loose body language. He has gotten a good look at Game, had the chance to sniff where she sniffed and where she stood to collect information – that’s all he needs.

02:46 Dog #5 is done; ready to head back home. He has learned all he needed/wanted to learn about Game.

02:56 Even when Game is back outside the forest, dog #5 is still good: he has satisfied his curiosity and is ready to return to whatever he was doing. (Probably snoozing outside his house.)

A glimpse into the life of village puppies: long version, and a detour on different types of selection

There’s a litter of four puppies in a 1000-habitant village in the State of Mexico. The day I made this video, I met two of the four. Only over the last couple of days had they started coming out and exploring: they had reached an age where they dared venture further and further from their birthplace.

It’s interesting to observe how many of the experiences Western breeders and puppy owners recreate happen naturally for a puppy like this – and they happen on the right time scale since it is the puppies themselves who decide when they are ready to explore, and how far they are ready to go on any given day.

You can also see differences within a litter: the two black puppies are bolder than the blonde one who is not with them, but who I saw the day after I took this video1, still in the safe space of the restaurant. The second blonde puppy must have also been a bold one – maybe the boldest one, or just a bold one with bad luck – because the person I am talking to in this video tells me that puppy got hit by a car earlier that same day.

The fact that within this litter, there are both bold and shy individuals shows an interesting tendency in evolution: evolutionarily speaking, both bold and shy individuals get selected for. We see this in humans, too. If a trait gets selected for, it has to have an advantage – and indeed, it does! It may seem counterintuitive, but in fact, both extremes of the spectrum of boldness and shyness can be advantageous. This is, I’d venture, particularly true for species that live in a vast variety of different environments – such as humans, canines and felines! Since the environments vary greatly, what is an advantage in one environment can be a disadvantage in another one. Or what is an advantage in one part of the year can be a disadvanrage in another part of the year. Or depending on what circumstances you happen to be born under – depending on random factors! – it may be advantageous to be either bold or shy.

A thought experiment: the shy puppy in the litter – the blonde one who I haven’t seen out in the street – is the least likely to get run over. From this point of view, being shy is adaptive – it increases the chances of survival, because cars (as illustrated by the death of the fourth puppy) are a HUGE danger to puppies. On the other hand, the two black puppies in this video show a lot of exploratory behavior, and they find food – both in the street and in the entrance of the store they then get shooed out of. From this point of view, being bold (showing a lot of exploratory behavior) is adaptive – it increases the chances of survival because you find more food. Due to studies done on puppy mortality, we know that most of these puppies are not going to survive. If one of them does survive – will it be a bold or a shy puppy? It could be either, because it depends on many factors: are the puppies still getting fed within the safe space of the restaurant? If so, being shy may be more advantageous because there is no lack of food resources. Are they not getting fed anymore now that they are a little bigger? If so, being bold might be an advantage because you need to learn to find enough food to make up for the calories you spend growing and existing! Being bold likely also increases a puppy’s chances of becoming an owned village dog, and owned village dogs get fed. If you are bold while you are still young and cute, you’ve got a killer combination setting you up for success in this respect … unless, of course, you get run over by a car first.

So there is no straightforward answer, but one thing is clear: depending on when, where, to whom and under what circumstances a puppy is born, boldness, shyness, or both may be advantageous. The same goes for humans. If it were not the case – if you were most likely to succeed by being a middle-of-the-road animal – the extremes of the spectrum of boldness and shyness would already have disappeared (for canines as well as humans). We would have what is called stabilizing selection: selection around a stable phenotype around a mean (a certain degree of not-too-bold-and-not-too-shyness). What we actually see is disruptive selection: selection at both ends of the normal curve: on the one hand, we get very bold individuals, and on the other hand, very shy ones. We see it in puppies, even within litters. And we certainly see it in humans, too! Even in very young toddlers, the differences are striking. By the way, a shout out to Marc Bekoff: I’ve learned the terms stabilitzing selection, disruptive selection and directive selection (selection for more or less of a given phenotype, e.g. if over time, puppies would tend to get bolder and bolder) from his book A Dog’s World, which I’ve had the honor of translating into German.

After this little detour into different kinds of selection, let’s get back to the experiences that breeders and puppy owners recreate, but that happen quite naturally for free-roaming puppies:

1. Introduction of different surfaces:

in the space of the restaurant, the puppies would have encountered artificial turf and real grass. Venturing out, they get to move up and down the stairs to the restaurant entrance, and they will walk on concrete and asphalt. In this video, one of the black puppies walks over an iron grid covering a drain – something else a breeder or owner might carefully introduce to their puppies that happens naturally in this environment.

2. Introduction to different sounds:

Currently, the 9 days leading up to a catholic holiday are being celebrated in this village – and like most Mexican celebrations, they are celebrated quite loudly, with lots of cohetes (firecrackers). Similarly, there are cars going by – this is the busiest part of town – and the puppies will get used to the sounds of cars, busses, motorcycles and lots of different human voices: adults talking and yelling, children laughing and playing …

3. People:

I’ve seen kids interact with the puppies (hold them, pet them, pick them up), and the puppies will also see people of all ages once they start venturing out of the restaurant space. People are quite naturally being paired with food, so a positive classical association is made to them when a puppy is born in the town center. They will also interact with people in that they get a basic village dog education: being cute and begging politely is going to get reinforced with food, and being obnoxious or entering forbidden spaces is going to be punished (at 09:18, the owner of the store across the street shoos the puppies back outside).

4. Dogs:

In this video alone, you’ll see three adult dogs: the fluffy dark dog, the pitbull, and the black lab mix. Throughout the day, the puppies will interact with A LOT of village dogs: everyone who roams freely, whether they are community dogs or owned free-roamers, will meet these puppies and interact with them. Some will be big, some small, some male, some female, most intact and some spayed. It is unlikely that a puppy born to a breeder would meet this many dogs at this age.

5. Other animals:

Sometimes, horseback riders come through; sometimes, they’ll see a cat, and once they are bold enough to venture just a little further up the street the store is in, they’ll see sheep and chickens.

6. Objects:

The restaurant is closed, but there are still chairs and tables in there. And once the puppies venture out, they’ll see cars, busses, and everything sold at the little stores around the area: brooms and food and buckets … At some point in this video, you’ll see one of the puppies approach a broom that’s for sale.

Further remarks:

+ I met the third blonde puppy the day after recording this video – so there must have been 4 originally, but 1 got run over, leaving three.

+ At some point in this video, I say that my AirBnB “tenant” also owns the restaurant – I meant to say host. I do not own a building in this town.

+ It’s interesting that I get asked whether I want to take the two puppies (they are community puppies, so unlike the puppies of owned village dogs, they are up for grabs). I assume the reason the person I’m talking to suggests I take them is that I’ve shown an unusual level of interest in the puppies – I’m following them around, filming and talking about them.


(1) One of the puppies is still alive for sure 2 months after I took this video, as I am writing this post – and it’s the blonde puppy (the shy one). I don’t know about the two black ones. They must either have died, or been taken in and have become owned village dogs. Statistically speaking (given the percentage of puppies that survive), they are more likely to not be around anymore – but we don’t know if this is the case for this particular litter. It’s a littler born under relatively advantageous circumstances, and in a good spot. (No highway; plenty of people; close to a food source.)

Distractions as cues, day 20: a naked dog win!

Naked breakfast:

Lovely response. I’m now getting what I originally hoped for! And a reminder that I can really keep my criteria for this learner tight: a week in a new location max. If I don’t see tangible results by then, even if I did in my previous location – time to change something!

It is interesting that this morning, just as in last night’s video with the collar only, Game approached the kibble further than in the previous reps with the rope. And that is even though today, the environment is same old, same old: no open gates or other unexpected changes.

It is still entirely possible that she cues off the kibble as soon as she notices it. It is also possible that the rope on her collar was a contextual cue for her to expect a distraction, making her more attentive and helping her notice the kibble pile earlier.

In any case, I am getting the desired behavior – and in fact, Game approaching the pile more closely is a nice opportunity to see what happens when I delay my click. So I’m going to stick with this project for one more session to find out.

I don’t think she will go for the kibble at this point. If she did, I’d say “Leave it,” and she would leave it. What I expect is more likely to happen is that she just stands and waits for further instructions, or that she starts coming back to me without the marker cue. I’ll find out tonight!

Naked Lunch1 dinner:

Since Game has been approaching the pile of kibble more closely lately, I decided to shape an auto-return ir auto-stop-and-look-at-me-and-wait before ending this project. So far, I’ve marked the moment of reorientation. Now, I’ll wait to mark her when she’s started moving towards me, or duration of her stopping beahvior – whichever she gives me.

Returning to me it was! Yay, Game! I planned to just take one or two steps, but seeing the confidence in her auto return, I let her take a few more before marking. That confidence showed me she was going to complete that return, so I could let her come closer before the click without creating insecurity about what I expected.

If I wanted to work on this project more, I would mark a few steps later on her auto-return in my next session. Then, I’d lower criteria in terms of what I expect of her, but make it more difficult by not stopping as I get around the corner, but:

  1. Walking in place
  2. Taking one more step towards the pile of kibble
  3. Taking two more steps
  4. Etc.

I would mark the moment of reorientation each time until I was confident Game was able to reorient when I was moving rather than standing still as well. Then, I’d build the duration of the auto-return back up. I won’t be doing this in practice though, at least not right now, because I have another training project I want to dedicate more attention (and treats) to. So for now, this will be the last day in this series! 20 days – not bad at all!

Also, meet the intermittent cat!

Fun intermittent cat fact: tortoise shells cats are usually female because in order to get that color, you need a gene for black fur on the X chromosome and an orange-spot-gene on another X chromosome (they cannot be on the same X chromosome, and neither of them can be on the Y chromosome), and XX will give you a female cat. You can technically have a male tortoise shell cat, but this would be rare, so because hoofbeats and horses, this one is likely female. In order to get a male tortoise shell cat, it would have to be an XXY cat (this is not lethal, but means the male cat will be infertile. It could also be a chimera, but I don’t know how exactly this would work. Thank you, Christian Holeček, for explaining the tortoise shell thing to me again after I forgot the details the first time I asked you!


(1) Should you read it? Yes, you should.

Distractions as cues, day 19: 1m, off leash!

Session 1: breakfast on a 1m rope:

Whee! Success on a 1m rope this morning!

I used to say I may shape a recall rather than just a reorientation once I have a naked dog. However – this may not be necessary because Game now stops and reorients as soon as she notices the kibble pile! She doesn’t approach it and then wait for me to mark, but stops very close to me, and pretty far from the pile. There may be no need to shape a recall if she keeps stopping so close to me!

Session 2: off-leash dinner

Off leash success! Game approached the kibble more closely this time, but I suspect it was not because there was no rope, but because the callejon looks different than usual (there’s a large gate behind my camera that is usually closed, but was open, making things look different). Game may have been distracted by/interested in this different picture and have noticed the kibble later than usual. So this isn’t something I worry about. I’ll be raising criteria to a naked dog tomorrow morning!

Distractions as cues, day 18: 3m, 2m …

Session 1, breakfast in location 2:

I cut about a meter off the rope, and got the same beautiful result. Will cut another meter by tonight. (Why? Dragging something may be a factor for Game, even if the weight of the biothane leash wasn’t. I will soon no longer be able to step on the leash, but Game gets to practice the familiar behavior with only a small increase in criteria.

My goal behavior here is to work up to a naked dog (she’s naked most of the time) – no collar. Then, if I feel like I’d like to take this further, I might shape her to come further back before my click by delaying the click more and more. Or I might end there – it depends on how I feel about my training schedule and how busy I am.

Session 2, dinner in location 2:

Success at 2 meters!

Btw, what Game is looking at here after finishing her kibble are the people talking on the patio ahead and to her left, not the cat.

If things keep going as well as they have been going, I’ll have an off-leash dog by tomorrow night!