What is the difference between disciplined inquiry and the scientific method?

Dr.Peter Sherrill's an entrepreneur and communications theorist. Welcome to Social Thought, Dr.Sherrill. We have a long tradition about what is empirical and what is scientific. Now, why are we so far off on the notion of empirical and theory?

PS: The confusion about the meaning of both of the terms. I mean, when we use the term empirical, it's used in many different ways, and when we use the term scientific, it's used in many different ways. And so, when you say there's confusion about being empirical and scientific, it's no question about that because we're not sure what we mean by science at any given moment, whether it's an activity by some people in a laboratory or whether it's a way of thinking, and empirical often means some sort of desperate cut-and-try methodology, and yet, in the language of analysis we would probably not use the term scientific, but substitute for it some notion like discipline, a disciplined way of looking at evidence, the kind of evidence that we look at with the observations of an empirical sort, that is, an attempt to make observations on the real world. And, in this tradition, it's interesting that a fact is simply a particular observation that many, many people agree to. The way to make the observation is described and, if a lot of people agree with what they see, then it's called a fact. But it's really just an observation that people agree to, made on the empirical world. And, in the same tradition, the empirical world is unknowable, and that's what gives Western thought at that point some mysterious and mystical idea which is not unlike mysterious and mystical notions in other cultures, or other philosophical systems. That is, the only world we deal with in what we consider a scientifically disciplined thought, is a world that's consistent with our ability to observe.


MP: A fact is traditionally treated as something standing on its own, and we have this amorphous, unknowable, empiric world, a fact somehow or other is a hard marble, it's, in the traditional view. Now, why is that such an inappropriate view?


PS: Well, there you have to get more rigorously into the language of analysis, where we have objects and we have attributes. And an object is part of is an empirical entity that is never knowable except through its attributes, that is, its characteristics and its aspects, its parameters, as they would say in engineering. So, by observing the attributes of objects, we come to infer the presence of an object, and the object is never known or understood or seen in its entirety, but only through its characteristics. Now, most objects are believed to have characteristics which may not be observable, which gives them a kind of mysterious quality. But it's the only the observable ones which are the ones that concern us and, as long as the observations are shared, the observations are shared on an object, we say that it's a fact. It's a fact that oranges are round only because many people agree with that observation. Our sensory apparatus tells us that they're round, and many people agree with that. And so it's a fact. But there are no other facts that can be assumed to not be tentative. And so, in the discipline tradition, almost all knowledge has a kind of tentative character about it, a speculative character about it. And then, to believe that there are ironclad realities and facts is a mistake which has been observed by many people.


MP: We're beginning to substitute the word discipline for science and I think that's highly appropriate, because you're suggesting that, as we begin to understand what the discipline is, we begin to have a more rigorous approach.


PS: The term for a while, people were using this disciplined inquiry instead of the scientific method because there is no scientific method, but there can be understood a general form of going about an inquiry that disciplined and follows certain rigorous rules. And, maybe in former times that would be called the scientific method, but I think any attempt to describe a scientific method, whether it's the hypothetic or deductive process or whatever, are all forms of disciplined inquiry, and that's much better to think in terms of discipline and inquiry rather than a scientific method for reaching truth.


MP: Before we get to the notion of an empirical test, I know we have to go through the steps of rigor, definition, as well as explication. What is that sequence before the empirical test has any ?rigor?


PS: Definition simply gives you the conventionalized meanings of terms, and an explication gives you an explanation of the idea. I tell you how I go about making my observations, the degree of rigor that I use in telling you how I go about making the observation is called an explication, that is, an explanation of my concept to you, and the rules that I give you are called operational definitions for making the observations. So, we can say, What do you mean by round and I can give you a set of descriptions and set of rules that will allow you to go out and look at something and say, Oh, yes, that's round. In slogan terms, years ago, people used to say, How do you know it when you see it and that's the operational definition. So we have now an explication, which is calculus for getting from an idea to an observation. And, the operational definitions are the final rules that you use to make an observation. And when you make the observation, whether you do it through instrumentation, electronic apparatus, or just with your eyes, however, it's still the operational definition. So it can have negatives. You can set aside conventional words. For example, in discussing the concept of poverty, we can say, well, poverty doesn't really mean lack of money, it means lack of opportunity, so we'll set aside money, it doesn't have anything to do with poverty. So, it is a way of refining the discourse, the meaning of the concept, but only for the discourse at hand and for a particular purpose. I don't think there's such a thing as a universal, global explication for anything. They all have this tentative character, and explication is used for a particular purpose, for a given discussion, a given domain of discourse. And its meaning is refined so that everyone understands exactly how it's used and how the observations are to be made. So, generally, in mechanistic terms, one would say, you describe all of the attributes that you can think of that are associated with a concept, you will set aside those that are not observable, and you concern yourself with those that are observable. And you then decide which of the ones that are observable most closely reach the center of the concept that you have in mind. Because there are observations that may not discriminate your concept from another one. The notion of human intelligence comes to mind here, that we can often mean many, many different things by intelligence, and the way that you observe intelligence can be done by what observing someone's verbal behavior, testing their mathematical behavior, making other kinds of observation of that individual in different circumstances. And each one of those may be no use at all. And very often in explication you end up throwing out the common for the concept that you had in mind. This is not uncommon in science at all, where terms have a very specific meaning when used in discourse in physics, for example, as opposed to in everyday language. Mass and gravity are terms which are used by everyday discourse, which are not used precisely, and yet, in science they have precise, very refined, specific definitions and meanings, and those meanings and definitions have come about either directly or indirectly through some process of refinement which we call explication. Oh, I suppose one could substitute explanation. I really like the idea of, explain to me what you mean rather than give me a definition for the terms. And the business of explication is concerned very much with explaining what you mean, what the ideas that you have in your head to me. It doesn't matter how many words it takes, as long as you finally communicate it. As opposed to going to a dictionary and pointing to the definition, which settles nothing and doesn't enlighten me, and doesn't excuse you of your obligation.


MP: How do we move back to the traditional concept of theory? I know that certainly in our generation, with Hempel and Kuhn, theory has become a focus of what disciplined inquiry is.


PS: Theories are proposed. Theories come out of human thought. Theories very seldom come out of empirical work. Theories arise out of speculative thinking about the way the world is. But theory is really a description. It's the most rigorous way to describe a theory would be to say two or more concepts joined grammatically. It's very tempting to fool with the numbers in economics and find ways to summarize things, look at variances and draw inferences, but we do have difficulty controlling the economy, which are, through this, which would require an understanding, and these models do allow us to make certain kinds of predictions, on the one hand. But generally, if you look in the behavioral sciences, you see that in psychoanalytic theory we have models of behavior that allow us to explain everything and don't allow us to predict very much. A good theory, that is a good system, a good conceptual scheme of the concepts are well explicated and are linked to each other properly. They both explain and predict, with equal measure. And a model which only can predict, without leading to understanding or explanation, is dissatisfying, or unsatisfying, and ultimately unsatisfactory in our ability to build on, build on that knowledge. And, and the, a model that allows us to explain everything without predicting anything is useful and useless in a practical sense. And the one that's invoked most often is Freudian psychoanalytic theory, which explains everything that's happened, but can't predict much of anything that will happen. And, the same is true of weather, the same is true in many areas of economics. A friend of mine once said that we build great linear models for handling the economy and we wiggle all of the x's and y never wiggles. And when we want to do something about inflation and we do all kinds of things to the independent variables and nothing ever happens to the dependent variable. We're not able to control it, because we don't understand, although we can put in more and more terms and ultimately the Fed can manipulate certain things having to do with the money supply that will bring about certain results, but not always, and not infallibly, and sometimes with unexpected and occasionally catastrophic results. The more clearly and more directly connected our system of ideas is, to the real world, the more we are able to predict events that take place in the real world by thinking about the concepts. Dr. Peter Sherrill, entrepreneur and communication theorist. I'll drop the orange for a moment and go to say oxygen, the theory of oxygen and combustion as opposed its predecessor, phlogiston, which was sort of an ethereal phenomenon that was involved in everything, it permeated everything. How are these two theories of combusion different?


PS: In anything we want to understand in the environment, in our attempt to understand it when we make observations, there's a great temptation to impute causes to the observations, and so when people watched fires burning, there was a natural tendency to understand why the fire burned, what made things burn. And prior to Lavoisier, of course, the classic example, that everything was held to hold some gas called phlogiston which was really burning and escaped and burned, and Lavoisier thought about that a long time and wondered why after something burned it was heavier than it was before it burned, and if phlogiston was escaping and burning, it should be lighter. And this led him to the notion that burning was really the combination of oxygen with the material which led to heavier, more mass. Phlogiston was simply a convenient concept, and when you invoke a convenient concept that hasn't been demonstrated with an empirical connection through a kind of formal explication or proper operational definition, is just loosely injected into a conceptual scheme to make things hang together and fit better, plausible explanation, they're called hypothetical constructs. And in social thought particularly, hypothetical constructs abound, there are very few well-explicated concepts. And in physics, we still have a lot of hypothetical constructs. And we have many that are held to be concepts that are well explicated and well defined, but they are vaguely defined. Gravity is one which doesn't have a particularly good definition in physics. So theˆ„, we must somehow, there's something in the way our brains work I think that forces us to inject an explanation, or to fit in a concept to make things hang together better. Perhaps it's aesthetic, it's an aesthetic need to complete the art of understanding. So maybe we should call, rather than disciplined inquiry, it's the art of understanding, which means that we construct loose ideas that are full of plausible but untrue constructs that make it all hang together. The business of disciplined inquiry is to get rid of those and reformulate them and purify them in order to clarify our understanding as well as the discourse.


MP: Derived from you decades as a survey research theorist in which we are dealing with statistical sampling process, and you've observed that statistical processes have come to dominate small-particle quantum physics. Now, how valid can a model, say quantum physics and chromodynamics, be if it's based on statistical observations and descriptions?


PS: I don't think that it can be very helpful at all. I think we're at a real dead end here. There's not much difference between quantum mechanics and statistical inference. And with statistical inference, we draw some inference about the nature of a large body of phenomena by sampling from it, and by understanding averages and the nature of the dispersion and the distributions or the departures from the average cases, in a kind of global way of understanding a large number of events. And certainly in thermodynamics, and certainly in physics, where there are a lot of particles, the idea to somehow summarize the actions and activities of the thermodynamic particles and the atomic particles, it's very tempting to find a mathematical tool that will allow us to do the summary without really understanding the process. And this is really something that Kurt Lewin pointed out, perhaps thirty or forty years ago, in the difference between what he called the Aristotelian and Galilean modes of thought. And Aristotelian modes of thought were concerned with average cases and average behavior, and with Galilean modes of thought we're concerned with pure cases and pure behavior. And if, he points out, that if Galileo had dropped a whole bunch of different boxes and things from Pisa, he probably would have been satisfied that on the average heavier objects fall faster than lighter ones and so on, which in fact is what he did not find out because he concerned himself with the pure case. If you were to concern yourself with experiments being performed in the real world where you concerned with wind and air resistance and all sorts of other things, the average case of Aristotle probably would have won, and we would have been led to different conclusions about the nature of falling bodies, is what I'm trying to say. The interesting thing to me is that there is such a thing as a pure case and there are also, as opposed to average-case analysis. And when the events that we confront are so complex, it's always tempting to use statistical solutions to allow us to say something useful, but it's only a summary. It does not contribute to our understanding. It only contributes to our ability to predict and summarize. And I think the same thing may be true in physics, that by our reliance on quantum tools, we are using summary procedures which are summarizing the behaviors of particles, rather than understanding the pure behavior of particle by particle.


MP: Supposing we had a pool table and, without any Newtonian mechanics, we just had the equations for the direction of the ball, the reaction, the incidence of a ball reflecting from the sides of the table, presumably with a modern computer we could predict very accurately where the ball would move, and yet, without Newtonian mechanics we would have no understanding of the forces involved. Where are we going from disciplined inquiry to a balanced model?


PS: We need to understand better I think what I mean by pure-case analysis as opposed to an average-case analysis. If we allow an event to repeat many, many times, as they do in economics, we can make certain inferences about what happens in the average number of cases. If someone is ill, a doctor can look at treatments, treatment modalities, and say, well, thirty percent of the people will be improved if I treat this condition, this pathology, in the following way, without ever understanding the disease or having no real knowledge of what's happening. And, and the pure-case analysis concerns itself with the detailed understanding of how, for example, a nicotine molecule causes certain impairments in the lung. The fact that so few people who smoke get lung cancer, ya know, leads you to believe that while smoking may be detrimental, we're still a long way from really understanding in the pure case the molecular processes that are involved. So the pure-case analysis is what needs to be dealt with here. What do we mean by a pure-case analysis? It means looking at only those concepts that are related in the process, are relevant to a process, and concerning our analysis with them only, just as Galileo did with falling bodies, just as Mendel did with his pea patch and the genes, so we have a different way of doing inquiry. And when the tools of averaging that are so powerful and allow us to make predictions, and deal in a certain way with the world, are not useful in advancing our understanding, and there's always the temptation, I think, to believe that our ability to predict, and to a certain extent manipulate, allows us understanding. But it is very limited, it's only confined to the specific, the specific condition for which we have the average. It doesn't lead to further understanding or any deeper understanding which will allow us to make improvements, in the environment. This I think probably is, personally for me, one of my greatest possible concerns is our failure, our rejection of pure-case analysis, because we have so many wonderful tools, and the computer of course aids and abets our ability to deal with average cases because it can do it so well. And pure-case analysis requires nothing more than a brain and a pad and pencil, generally.


MP: I want to give the listener an example and I want you to correct me. This example would be where we're trying a new drug against some control situation, and we have a large sample. As I interpret what you're suggesting is that we apply the drug, we look at the results, and then we use various types of averages, factor analysis, a range of averaging procedures, and then we go and apply it to what we consider to be the universe, later on, after we have the drug results. What you're suggesting is that, when we look at the results at the first step, that we don't average, that we go and look and we say, for this group of ten people, it had this effect. For this group of twenty-two people with these characteristics, it had another effect. And that we look at the cases where the effects and the drug made some sense, was understandable, and we simply transmit that straight through to the final treatment, where, if the patient has the qualities of this twenty-two and the effect, that's what we're looking for.


PS: Yeah, that 's lovely, Michael, because what you're saying is, and what I'm trying to say and didn't say it as well.