Jon Katz has an excerpt from a new book of his on Slate, asking whether dogs can think. Okay, possibly an interesting question...I wouldn't ask "Do dogs think?" (I think that's a given), but "How do dogs think?" But anyway...

He starts out with an anecdote about a woman who goes back to work after a few years of being at home. Her dog's behavior changes for the worse at this point, chewing up furniture and whatnot.

"I know exactly what's going on," Heather told her vet when she called seeking help. "Blue is angry with me for leaving her alone. She's punishing me. She always looks guilty when I come home, so she knows she's been bad. She knows she shouldn't be doing those things."

Heather's assessment was typical of many dog owners' diagnoses of behavioral problems. And her vet agreed, suggesting "separation anxiety" and prescribing anti-anxiety medication for Blue. Heather also hired a trainer, who confirmed the diagnosis.

Katz then goes on to say that attributing human emotions and motives to the dog is wrong. Well, okay, he's got a point. So he calls up a canine behaviorist at Cornell and gets the "real diagnosis":

"Being angry at the human and behaving punitively—that's not a thought sequence even remotely possible, given a dog's brain. The likely scenario is that the dog is simply frightened." When Heather was home, she was there to explain and enforce the rules. With her gone, the dog literally didn't know how to behave. The dog should have been acclimated to a crate or room and confined more, not less, until she got used to her new independence.

First of all, the first assertion, that being angry at a human and behaving punitively not being remotely possible...I'd like to know what he's basing this on. It may not be the case in the specific example...but is he saying dogs don't experience anger? Or that they don't attribute anger to people or other agents? Or that they don't retaliate against agents who are the cause of their anger? I might doubt how elaborate this thought process is in a dog, but I certainly wouldn't entirely rule it out.

Then Katz says something that completely undermines pretty much any credibility he had to begin with:

Dogs are not aware of time, even as a concept, so Blue couldn't know whether she was being left for five minutes or five hours, or how that compared to being left for a movie two weeks earlier.

Now this just smacks of downright ignorance. Of course dogs are aware of time, at some level. Hell, even fruit flies have an internal clock. Innately knowing what part of the day cycle it is, and keeping track of the passage of time, is important, if not crucial, for many organisms.

Much of our memory is episodic. That is, we remember events, or sequences. Remember the time that X then Y then Z happened, in that order. All the evidence that I've read indicates that other mammals, and possibly down to some invertebrates, store episodes rather than static, non-temporal instances.

I was talking to another student last week about a specific kind of cognitive task called a Delayed Response Task, which tests working memory. The basic idea is that two cues are presented to the subject, like a red or blue light. Then there is a delay phase, usually seconds long. Then there is a trigger (e.g. a green light). The subject has a choice of two motor responses, and if it associates the correct response with the correct stimulus, it gets a reward.

One example is a T-maze task. A mouse is put in a maze shaped like a T, so it can move forward and then go either left or right. At time 0, you present one of two sounds, a ding or a buzz. Then there is a delay interval. During this time, if the mouse moves forward, it gets no reward...it has to learn to wait. After the delay period, you signal with a tone, the trigger. If you want the mouse to associate the ding with going right and the buzz with going left, when it correctly makes these associations, you give it a piece of cheese.

Okay, so the mouse has some sense of time. This is ridiculously obvious. An elaboration on this is to not have the trigger, but to only reward the mouse if it waits a certain interval before making a decision. You could increase these intervals to see how long the mouse could internally process the passage of time. You could also devise studies that test long-term memory.

Hell, I've read of insects exhibiting long-term memory. There are wasps that bury their eggs with fresh caterpillars in several different locations, and only visit those egg sites every 2-3 days to replenish the food supply. At some level the insect is internally tracking the passage of time.

You could get into a philosophical discussion about whether the wasp or mouse or dog is processing time more like we are or like the computer you're reading this on are. That would be an interesting discussion. Because it's ultimately subjective, it would be hard to know exactly.

But it seems horribly irresponsible to rule out the idea that dogs process time in any way that resembles the way we do.

Katz's article warns about the danger of attributing thought process to dogs that are too much like ours. But he goes too far in the other direction, I think, classifying the dog as a purely instinct-driven entity, with little or no overlap in the way they think and the way we think.