Investigations of logical thinking

Dogs derive from the wolf. Though we know that in the course o domestication they haved adapted to the living together with man in various aspects, its not yet clear whether their thinking still functions "wolf-like" or whether their behaviour in solving problems as well as their learning ability have changed.

One hypothesis supposes, that dogs in the course of domestication have lost somewhat of their cognitive intelligence. A basic assumption thereby is, that man die solve a lot of problems for them, while wolves had to cope with the challenges of the wilderness on themselves. On the contrary a second hypothesis assumes, that dogs did not loose any of their causal, physical comprehension, but became mor sensitive to human influences and thus their behaviour more often is controlled by human behaviour than by independent conclusions from physical causalities.   

At the Wolf Science Center we try to investigate these two hypotheses more exactly.The touchscreen is a super method therefore. Using it we can test logical thinking and individual learning without influence of man – rof the animals works at the computer and not with us!

n 2007 we started in the Clever Dog Lab , to use an automated Computer-Touchscreen-Procedure, developed by us, to test the behaviour of learning and the locigal thinking of dogs. Now we use the same procedure with our wolves at the Wolf Science Center. It functions like that:

During one testing sequence the dogs have to prod a picture at the screen with the nose, whichone is defined to be positive (meaning the animal gets a reward). After a phase of accustoming (familiarization) they are trained to visual discrimination tasks:

  • First the dogs have to learn to chose one out of two simultaneously presented pictures. We  use simple forms (squares and circles).
  • In a second task they must learn to look mor exactly to discriminate two sets of stimuli. One set shows gay coloured underwater pictures, the second one drrawings.

After the animals successfully have mastered both tasts we start asking more difficult questions, e.g. whether wolves and dogs can learn by using the „exclusion principle“  – that's already a matter of locigal thinking.

All of our animals are true computerfreaks and would sit for hours in front of the "idiot box" being allowed to!Our dream is an automated Touchscreen-station, which can be visited by the animals to "play" at any time – but we are still looking for a generous sponsor financing it.

Learning by the exclusion principle

To learn by the exclusion principle means, that one choses a correct alternative by logical exclusion of a potential option.

Explanation board with examples of the symbols used in the exclusion principle Photo: Clever Dog Lab
Exclusion principle

In man the ability to learn by the exclusion principle is well developed  (e.g. when learning new words during language acquisition). A couple of studies showed, that also some species of animals can learn by the exclusion principle, whereas others are not able to do so. Since these results eventually are due to different experimental methods, we investigated learning by exclusion principle in pigeons (n = 6), dogs (n = 6), children (n = 8) and students (n = 6) using the same method for all. Thereby all of the subject group had to learn first, which of two simultaneously on a monitor shown pictures of all-day objects either is marked positive (S+) or negativ (S-). Alltogether there were four positive and four negative pictures shown in all possible combinations.

The object (picture) was chosen by touching the screen with beak or nose (pigeons, dogs) and by a mouse click from humans.The correct choice (S+) was rewarded, an incorrect choice(S-) was followed by break and afterwards the repetition of the choice concerned. After the subject group was able to reliably discriminate the objects, they were confronted with test, now the monitor presenting only one of the negative marked objects together with one of  four new objects(S. One pigeon, three dogs and nearly all humans chose the new object in each case. This can be due to the preference of the new against some familiar (old) or to the avoiding of the negative object without any learning about the new one.

This choice can also be achieved using the exclusion principle, in which the "new" object (S') now is classified as positive, since the alternative is negative (S-), as one knows. To decide which mechanism of learning was used by the subject group, the participants with a preference for the "new" object now were confronted with tests, wherein the "formerly new" objects (S') were presented together with "complete new" objects (S'').

While the pigeon chose the "complete new" objects (S''), sticked the dogs as well as the humans to their preference for the "formerly new" objects (S') and thus showed that the negative marked object (S-) in the first test was logically excluded and the "new" object (S+) consecutively was defined to be postitive.

Aust, U., Range, F., Steurer, M. & Huber, L. (2008) Inferential reasoning by exclusion: A comparative study of pigeons, dogs, and humans. Animal Cognition, 11: 587-597.