According to the professor, what assumption might a reductionist make when analyzing ant colonies?

Listen to part of a lecture in a philosophy class.
Professor: Today let’s explore two important philosophical concepts. They are two basically different ways of looking at things that have wide ranging influence on many disciplines like psychology, sociology, and animal behavior, and these approaches actually guide the way scientists do their research. You might have guessed I’m referring to holism and reductionism.
    That’s holism—focusing on how an entity functions as a unit—and reductionism—reducing the whole to its parts. Suppose you’re looking at an ant colony, and the ant colony would be the whole, the whole system as it were and the ants are the part, with each individual ant representing the smallest part of that system. Reductionism is the belief that the workings of the system, any system, can be understood through an understanding of its smallest parts, how those parts work individually and also, the way any particular part interacts with another part. So in the case of the ant colony, a reductionist would investigate the behavior of individual ants within the colony and the way they interact with other ants in that colony. From that they would make generalizations about the behavior of the colony as a whole. Let’s take food for example. When an ant finds food, it takes that food back to the nest, and on its way back, it secretes a chemical, which leaves a trail that can be detected by other ants. In this way, it communicates about the location of food, and this behavior..., this communication describes an important way in which ants in a colony interact. To the extent we can understand the rule governing this behavior of that smallest part, the individual, then we can generalize about the behavior of larger groups of ants: following trails, finding food, leaving more trails, and so on. Eventually, we’ll understand the behavior of the entire colony and then, and this is an important part of the reductionist approach, we can predict what will happen in that colony.
    Now, on the other hand, a holist believes that a system cannot really be understood by just understanding how the parts function. In other words, an old saying you might have heard before, "the whole is more than the sum of its parts." So when looking at an ant colony, the holist sees it from two perspectives: of the individual ant, and of the whole colony, because the behavior of the individual ants taken together has an impact on the colony as a whole, and the colony as a total system has an impact on the individual ants. You have to view the behavior of the colony as more than simply the total of the individual behaviors. One reason for looking at the behavior of the colony is that the colony as a whole seems to know what to do, to act for its own benefit in a way that any given individuals could not.
    The holist believes that the collective behavior of the colony as a whole cannot be traced to specific individuals. In animal behavior, we refer to this as "collective intelligence," and it’s a good thing that ants have this collective intelligence because the individual ant is ill-suited to life outside the colony. So what’s an example of this collective intelligence? Well, sometimes the ants switch tasks, let’s say from taking care of the next generation to gathering food, and we don’t know exactly what triggers this behavior, but somehow they just know when more food gatherers are needed. There’s no leader in an ant colony, no one giving orders. And don’t confuse ant society with human society: there is a queen, but all she does is lay eggs. And of course, no single ant could possibly know how much food is needed in the whole colony, but somehow, as a group, they know to supply the colony with enough food. The most likely explanation for this is that it has something to do with chemical signals from ants in the colony.