Can wasps make comparisons like humans?

Human beings are very talented at using a cognitive skill called “transitive inference” - using information about things you know to draw conclusions about things you don’t know. For example, if you know that A is bigger than B, and B is bigger than C, you can tell that A is bigger than C without having to look at them side by side. We know humans can do this, but it’s something of an open question which other animals can do it as well. Professor Elizabeth Tibbetts from the University of Michigan has been investigating whether one of mankind’s most maligned foes, the wasp, is capable of using this advanced cognitive technique. She spoke to Ben McAllister...

Elizabeth - Long ago people thought that transitive inference was based on logical reasoning and we thought only humans were capable of transitive inference and, not too surprisingly, before long we found that humans were not the only ones. It turns out that a huge range of vertebrates can do transitive inference so primates and birds and even fish.

Ben - Wow. So it lives in that bucket of things that we used to think were kind of unique to the human experience, but we are rapidly learning is becoming a much much smaller bucket?

Elizabeth - It's a very small bucket I think. There had been one study on transitive inference in a non-vertebrate and that was done in honeybees, and they found that bees couldn't do transitive inference. And so I think that wasps are way smarter than bees so I wanted to test whether wasps could do it.

Ben - And for anyone out there who isn't super fond of wasps, you heard it here first, wasps are not only scarier than bees they are indeed infinitely more cunning as well, so add that to your consideration. What did you do with this study specifically in order to figure out if wasps could use transitive inference?

Elizabeth - What we did is we trained them to a bunch of colours. So, for example, we would train them that blue was better than green, and then we would train them the green is better than purple, and then we would train them that purple was better than yellow. So they had all this information and now we asked them to make an inference, so asked them what you like better green or yellow?

Ben - Right. And they've never seen green or yellow together before?

Elizabeth - Exactly. They've never seen green or yellow together. Some of the time green has been good, some of the time green has been bad so there's nothing that should be inherently different about the stimuli.

Ben - How do you go about training a wasp that green is better than say any other colour?

Elizabeth - We train them in this tiny little maize. It has to be tiny because wasps are tiny. Some of the bottom is electrified and then some of the bottom isn't. So when we're training them that blue is better than green, blue is a safe area in the maze and green gives them a little electric shock.

Ben - How exactly do you figure out what's a little electric shock for a wasp?

Elizabeth - I would say it's trial and error. But I promise no wasps were harmed in this experiment. We want them to learn so we don't want them to be freaked out or really worried or anything, right. So we just give them enough shock so that they act a little uncomfortable so they start moving around more quickly and try to get away from it.

Ben - And so they just spend a bit of time in this maze until they eventually land on the part of it that doesn't shock them and that part corresponds to the colour that you want to train them is good?

Elizabeth - Exactly. They move around the maze and they eventually go to the part that's safe and they're like oh my gosh, it's safe and there's the colour green - green is great.

Ben - Okay. What did you do after you trained them?

Elizabeth - After we trained them we had to test them, so we put them in the middle of a box and then we tested which colour they prefer to go to.

Ben - And you had no electric stimuli or was that still present?

Elizabeth - There were colours on either end and there was no electricity to cue them, but the idea is that they've learned that green is good, then they would go to the green side. So we tested them on the colours we had originally trained them on just to confirm that they had learned what we trained them on, and then we also tested them on those new transitive pairs.

Ben - Okay. And what did you find?

Elizabeth - We found that wasps do have transitive inference. So they took all those trained pairs and they seem to kind of organise them in their mind linearly, and then use transitive inference to choose between stimuli that had never been next to each other before.

Ben - That's fascinating because, as you mentioned before, previously someone else had found that bees are incapable of doing this. Surely a bee and a wasp have pretty similar sized brains, right?

Elizabeth - Yeah. Bees and wasps both have similar sized brains and their brains are really tiny, about the size of a grain of rice. I think the difference between bees and wasps isn't really that wasps are just geniuses and bees are dumb, it's more about what the social life of wasps and bees are like. All the workers on a bee colony are about the same, they spend their time foraging, but on a wasp colony there's all sorts of interesting dominance relationships. They have linear dominance hierarchy where the dominant wasp does most of the reproduction and the subordinate wasps do most of the work, and so figuring out how dominant other wasps are in wasp land is incredibly important. For example, if you've beaten Jane in a fight before and you see Jane beat Susan then you can infer hey, I'm probably going to be able to beat Susan. So that kind of thing is really important for wasps and not important for bees.

Ben - I would also say probably important for humans depending on who you ask.

Elizabeth - Yeah, definitely important for humans.

Ben - That's an important thing to know. Do you think there's scope for extending this kind of reasoning to dealing with other animals or other kinds of animal cognition?

Elizabeth - I bet that many other insects are capable of transitive inference. I think we just haven't tested them yet. I think one of the messages is that animals can be really good at what's important to them. We think of humans as being like the best at everything, but lots of animals are amazing at really specific things they need to do to be successful.

Ben - Intelligence doesn't necessarily correlate just the size of the brain but also to the tasks that need to be undertaken?

Elizabeth - Exactly. You don't need a big brain to do complicated things. Even a tiny little brain can do complicated things if the animal needs to be able to do it.