Named objects (building a vocabulary)

One specific object, and a name that belongs to it alone.

• Pick an object that is easy to recognise, clearly different from everything else in shape, size or texture, and safe for him.
• Give it a short, punchy name, distinct from the cues he already knows.

Move on when: Your dog already retrieves this object reliably.

The word and the thing merge into one.

• Put the object on the floor, two or three steps away from the pair of you, clearly visible and alone in that spot, with no other object nearby.
• Point to it or look at it, say its name in a clear voice, then let your dog turn towards it, take it in his mouth and bring it back to you.
• Mark with a word and reward, a treat given by hand or his favourite toy, as soon as he turns towards the object, then a second time when he takes it in his mouth: this double marking is only for the very first repetitions, to show him the whole sequence. As soon as it flows, mark only once, when he picks the object up.
• Repeat this sequence many times, over several days: the word really has to "stick" to the object before you move on to the next test.

Move on when: As soon as you say the name on its own, without pointing at the object or indicating it in any way, while it is still in the same spot, your dog turns decisively towards it, and this happens across several sessions in a row.

Choosing the right object, not the nearest one.

• Before you put the objects down, ask your dog for a sit-stay, or hold him by the collar: he must not rush at them while you are setting them out. Next to the named object, place a second object your dog has never retrieved and never heard named, neutral and with no play value in his eyes, about a metre from the first. Stay two or three steps away from both objects yourself, never between them and never leaning towards one of them.
• Release your dog and ask for the object by its name, staying neutral: don't look at the object you are expecting, don't turn your body towards it, keep your arms by your sides. Your dog reads you very easily, and a single glance or a step towards the object can hand him the answer (this is the Clever Hans effect).
• If he chooses right, mark and reward; if he gets it wrong, simply let the mistake go, never say "no", and try again with an easier set-up: move the two objects further apart, or swap in a neutral object that is even more different from the first.
• Change where the two objects sit on every attempt, left or right, closer or further apart: your dog must never be able to guess from the position, only from the word.

Move on when: Across several attempts where you have genuinely varied the positions, he clearly gets it right, roughly five times out of six.

Recognising the right name out of two, without guessing.

• Teach this second object exactly like the first, going back through stages 1 to 3, with the first object out of the picture.
• Once this second name is solid, present both objects together and ask for them one after the other, in a random order, changing their positions each time.

Move on when: He picks the right object out of the two, in a random order, before you add a third name.

A vocabulary that grows, without the words blurring into each other.

• For each new name, start again from stage 1 with the matching object: this is not going backwards, it is the normal rhythm, and the vocabulary grows slowly, one name at a time.
• Start with objects that are clearly different from one another, then move gradually towards objects that look alike, two toys of the same shape but different colours for instance: that is where the sorting really becomes precise.

Move on when: Every new name goes through the same success thresholds as the ones before it, before you add another.