When we talk about making guesses, we talk about the way that everything we can call "living" stays alive. Everything we call "living" stays alive in pretty much the "same" way.
 
That means that this discussion will take a very broad view of our topic, but it does not mean that the topic doesn't apply to humans just as much as it does to amoeba or algae.
 
 
First premise:
 
Everything living makes "guesses" about what it needs to live and how and where to obtain it. Of course I don't mean that algae or amoeba talk about this; I want to talk about "guesses" that don't have to use language at all. When you want a drink of water, you don't talk to yourself about where to find a glass, or how far to reach for the water tap; you do it. If you knock the glass over instead of picking it up, then your physical guess didn't work. You may say something at that point, but what you do corresponds to guessing again: you extend your arm at a slightly different angle, and move your fingers a little differently, and pick up the glass. What you said probably came out as an exclamation, but not as "extensor muscles do this and flexor muscles do that."
 
Why do we talk about "guesses"?
 
To start with, we claim that nobody and no thing "knows" for sure what goes on in and around itself. At its very best, it must pay attention to its sensory organs and what they tell it, both about its internal environment—what it needs to stay alive—and about its external environment, where it will look for ways to satisfy those needs. So on both an "internal" and an "external" level, the organism must guess about its needs and what it may find to satisfy them.
 
If the guesses work out—if the organism fills its needs—it will survive for at least a little while longer. Now this looks pretty simple when we speak of algae and amoeba, and even when we discuss fish and spiders and foxes. Obviously, if the environment they live in doesn't provide them with food, water, and suitable places to live, these organisms will die. Likewise, if the other, larger predators in their food chains feel hungry and snack on them, they will die. We can say with some confidence that they live by making the best guesses they can, and go on living if and only if these guesses don't get disconfirmed—or to put it in positive terms, as long as testing those guesses, which they do by living them—works out so the outcome satisfies the organism's needs.
 
Clearly, these organisms—birds and fish and foxes—don't talk to each other about their guesses; and we have no reason to believe that they talk to themselves ("think") about them, either. They test them. They act to satisfy the needs that their senses "tell" them about: "Hungry!" or "Cold!" or "Get ready to have babies!"—and if the outcome satisfies their needs, then that guess worked that time.
 
What do we mean, "testing in action"?
 
In our society, we usually think of "tests" in two ways. One looks like the tests we took in school—where we either had to choose the "right" answer or write an essay that explained something we had "learned." The other looks like the kind of test that scientists do in a laboratory, where they take some substance and do things to it to decide just what it consists of, or to measure some of its "properties", or show that it does certain things under certain conditions. To do any of these kinds of tests we must use language as well as some limited forms of action. These kinds of tests do not fulfill our criterion of "testing in action" for two reasons: first, they involve lots of languaging and "thinking about" and second, they set out to get a "right" answer, to "prove" something.
 
You may well ask at this point, "then why test at all? Don't we want the right answer?" No.
 
NO? No. We want an answer that works, here and now, in this situation, to take care of the need of the moment. A "right" answer would work always and forever, in any situation. But we don't ever get that kind of an answer!
 
NO? No. We live in a constantly changing world, and we produce physical needs that change all the time, too. If we rely on right answers, we don't take any of those changes into account. And we could apply that right answer at a wrong time or in a wrong situation. Remember, we don't mean answers on a test sheet now. We mean choosing actions to stay alive.
 
A baby turtle makes guesses as it emerges from its egg high up on the beach, and must scramble for the ocean. If it heads in the wrong direction, it will die. If a seagull dives upon it, it will die. If it reaches the ocean, but gets eaten by a predator, it will die. The baby turtle acts on its guesses as well as it can, and if all of its guesses work out, it lives. No guarantees—it tests its guesses by acting on them, and we know that these guesses do work when we see young turtles growing up.
 
A human baby makes guesses as it emerges from the mother. This infant has felt squeezed and pushed around and then dumped into a much less friendly environment than the one it has lived in for months, but if it acts sullen and resentful and fails to make appropriate guesses about what to do next, it will either die—or, given a good care provider, will get smacked, and take that first breath of air, and howl. The infant tests its guesses by breathing on its own for the very first time. We don't usually say that the infant "tested its guess" about how to live in a world of air—but that clearly describes just what it did.
 
 
Second premise:
 
Nothing stays "the same" from one moment to the next. Nothing at all. If we can count on anything in this world, we can count on things changing. Some things change quickly and some change more slowly, but everything does change.
 
This means that we have to make guesses. Just because we tested one guess in action a "while ago" and it worked, does not mean that "the same" action will work "now." If we try it, and it does not work, we need to guess again. We usually can do so, because many of our guesses do not have life-threatening consequences, or at least not consequences that threaten us immediately with death. Some do, of course. If we pick and eat the mushroom that looks "just like" the ones we picked the other day, we may find out by dying that we have guessed wrong. But if we just overeat—make a wrong guess about how much food we need, because a few years ago we could (and did) eat that much to satisfy the needs of a growing body—we may feel overfull, or get indigestion, or gain too much weight. None of these consequences will prove fatal, or at least not immediately.
 
If I could safely climb a certain tree years ago, why can't I do it today? But perhaps the tree has suffered damage, or perhaps I have grown heavier. Perhaps I don't climb as easily as I did then. If I "know" I can climb the tree just because I did so once, I will climb the tree. I need to remember before I start up, however, that neither the tree nor my "self" has remained unchanged over time, even over a few minutes. I may guess I can do it, and test that guess by climbing; or I may not.
 
 
Third premise:
 
As we grow up, we humans give ourselves more problems about guesses than foxes or infants seem to. We often use our language to interfere with what our senses tell us about what goes on in and around ourselves and also about what goes on outside us. We forget that each moment comes as a new experience; we pretend that we "know" what goes on.
 
Even our language encourages us to make these mistakes. For thousands of years, our culture has acted as though it really knows how things really work. As we assimilate our culture in the process of growing up—as we do the time-binding that make us "human"—we acquire this attitude. Every time we say, "This is the way to do that," or even "This is mine," we reinforce that mistake. We act as though we really know.
 
Worse yet, as we learn about the world, we get the idea that things stay pretty much the same over time. Even though our sciences keep telling us otherwise, our culture encourages us to rely on "old" information to get by. Worst of all, this works most of the time. When it fails to work, we feel indignant. Our world has let us down! The tree limb broke, the mushroom poisoned us, the baby we so confidently expected died at birth. In other words, our guesses failed to work this time. And instead of admitting that we had made guesses, and they didn't work, we look for someone or something to blame.
 
Blaming does not help us make better guesses; it insulates us from realizing that we do make guesses. So when we look for someone or something to blame, we avoid leaning the crucial fact about guesses: we have to test them in action; sometimes they work, and sometimes they don't, but if they don't we need to throw them out and guess again.
 
Guesses—tested in action—allow us to take one step at a time, and to keep on if and only if what we do seems to work: if the outcome satisfies our needs. We don't commit ourselves to an action just because it "worked at one time." We don't commit ourselves to an action because we "think it would work" or because it would "prove something to someone." Especially, we don't repeat an action that we have already disconfirmed because we expect to "do it better this time." We also remember that only "I" can make "my" guesses, and only "I" can test them. Even if the outcome of this guess works really well for "me" here-now, it may not work for anyone else, there-then.
 
Radical uncertainty
 
This means living in radical uncertainty. Not a pleasant or "safe" prospect, we may say. But it does describe (as nearly as words can describe) how all living things in this world do stay alive. No species that we know of other than the human pretends differently, and only we humans have to learn to avoid that mistake.
 
Martha Bartter