What is that little doggy thinking?

Everybody's dog "knows exactly what you're saying" - even, it can sometimes seem, exactly what you're thinking. In the mutual devotion of dog and human pack-leader, the rapport of daily ritual and association seems bound, from time to time, to prompt something very like telepathy.

Ethna has taken up the habit (or, at least, the aspiration) of a brisk two-mile walk on rising, thus sparking joyous hysteria in Meg, our young labrador-spaniel.

One morning recently, the dog and I were abroad in the garden early, leaving Ethna still asleep, curtains drawn. Half-way through our slow inspection of the vegetable beds, attended by rustling leaves and birdsong, Meg suddenly sprang alert and ran excitedly up the path towards the house, looking back to insist that I follow. There were two doors to open before we found Ethna getting out of bed. All she had done before that was open her eyes.

There followed a gap in the early walks: a pulled tendon called for almost a week of rest. Then came the morning when Ethna felt ready to march out again. Having done nothing more than think this, she emerged in pyjamas to find Meg dancing round, yelping in delight.

Unravelling such anecdotes for their precise sequence of unconscious human cue and canine association is next to impossible. Even with Big Brother cameras on the wall, what we know happened might be quite different from what Meg knows happened. If we sometimes think she thinks she's human, it's certain she believes us to be dogs, of even more uncertain breed than her own.

In theory, her apparent intuitions have their roots in the associative learning that so many of her daily actions make clear. The ideal dog's life, it seems, is all eventful comings and goings, galvanised by the jingle of car keys or a word processor signing off. Any sentence we speak that has "go" in it is likely to lift her head and light her eyes.

Even the "uncanny" perceptions of dogs seem often to be bound up with movement: the dog that "knows" that a family holiday is planned, even before the suitcases come down; "knows" when a family member has left the office; "knows" when the family car is nearing home. Rupert Sheldrake, whose books on the workings of "morphic resonance" have so annoyed fellow scientists, sees such behaviour as ready-made for do-it-yourself research (see his Seven Experiments that Could Change the World, published in 1994).

Mainstream science, however, now steers clear of these haphazard canine conundrums, to which the answer might end up as: "so what?" Ever since Pavlov conditioned dogs to salivate when a food bell rang, behavioural science has insisted that animal behaviour can be explained only in terms of stimuli and responses; any research into animal consciousness, thinking or subjective feeling must be inherently "unscientific".

By such criteria, all non-human species appear as simpler organisms, the level of their intelligence established by solving problems which we - not necessarily they - find significant. Consciousness is grudgingly accorded to those few animals that can be shown to possess self-awareness.

The classic test for this has been an animal's reaction to watching itself in a mirror. By much the same impulse that has a territorial cock-bird attacking its own reflection in a window, many animals, including dogs, react as if towards another of their species, then look around the back of the mirror, and eventually decide to ignore the whole affair. Meg's excited response to her own image in home videos has all the appearance of self-recognition, but other interpretations, I suppose, cannot be ruled out.

For chimpanzees, however, the mirror test has won a quite new significance. In what is presented as a clinching refinement, they will dab at a brightly coloured, red spot on their forehead, surreptitiously put there by the experimenter, and then, in some reactions, sniff their fingers.

Such tests are old hat to Sheba, the 29-year-old star of a British TV programme to be broadcast on Channel 4 next month. As The Cleverest Ape in the World, or at least in Sally Boysen's chimp centre at Ohio State University, she and her fellow apes will be seen engaging in feats of mathematics, linguistic comprehension, spatial awareness and problem-solving that dazzle the onlooker and while away the boredom of the animals' laboratory lives.

Large-brained apes and dolphins are now well-established subjects of such research, even if still unable to say what they really think about it. Some small-brained birds, on the other hand, can be trained to communicate in English. A new book from Harvard University Press is The Alex Studies: Cognitive and Communicative Abilities of Grey Parrots, which reports on a scientifically structured, 20-year love affair between trainer Irene Pepperberg and an African pet-shop parrot.

The bird's language remains the mimicry of human speech that its species (and others, such as starlings and crows) has always possessed. But in her use of "social modelling theory" Pepperberg has trained Alex to interact with her in tests of cognitive ability, using words that let him mean what he is saying.

Her work has been applauded, but her stated aims are modest: "to encourage an awareness of, and a sensitivity to the abilities of non-humans, particularly non-primate, non-mammalian subjects."

Donald Griffin, the Harvard biologist who discovered the uses of bat sonar, is sure that many animals give conscious meaning to their communication and behaviour. In his book Animal Minds, he holds a door open even to a consciousness of honey bees.

In my own love affair with Meg, pursued in nightly snogging on the sofa, there is much protracted, eye-to-eye communing at close quarters. Behaviourists would assure me that this is mere submission on her part, like the ritual licking of my hand and the grovelling abandon of her posture on my knee. It doesn't stop me wondering what she's thinking.

Michael Viney welcomes observations sent to him at Thallabawn, Carrowniskey PO, Westport, Co Mayo. E-mail: viney@anu.ie