16 Summary and conclusion
Wrapping up what is and isn’t scientific, what knowledge is certain, what we could hold more lightly, and how these concepts could be beneficial
This is the last post in this series. You may want to read the Introduction titled The Myth of Scientific Uncertainty, and posts numbered 1-15 first. under “archives.”
Here are summaries of the principal epistemological points of the earlier sections.
Theories have two distinct parts: laws and explanations. They are the “what” and the “why” of our knowledge. The laws, based on observations, do the quantitative and logical work, but they do not reveal why. That is the task of explanations which are our rationalizations, models, and pictures of why this is so. Their qualities and contributions to scientific knowledge are distinctly different, but equally essential.
Laws can be certainties, but not universalities. They have proved to work consistently within the limits of the conditions in which they have been tested. Applications beyond those limits are assumptions and associated conclusions are therefore provisional.
Limits to a law’s certain applicability are those conditions which haven’t been verified or they are conditions under which other related phenomena affect the result. Most laws have limits on the range of conditions within which they accurately apply.
Exceptions to a law’s applicability are limits on its application: they are not disqualifications. Having verified the certainty of a law within its limits, exceptions must occur under conditions outside those limits.
Catastrophic changes in conditions have only philosophical significance. Karl Popper names the disintegration of the world as a situation in which current scientific knowledge would fail to apply. Other situations would include our planet’s being swallowed by the expanding sun. While these are philosophically valid exceptions, in practical terms, there would be no one left to care (or gloat). In the meantime, the laws will continue to work within their verified limits.
Instrumental measurements are valid observations. Instruments of science are based on verified laws, which give certainty to their results when used correctly and within limits. Errors in their application show up in attempts to confirm or verify reported results.
The influence of unconsidered factors in the application of laws are determinable. Limits of measurement precision due to uncontrolled variables do not invalidate laws or make them “partially true.” Determining the variance in observations is integral to science and the results are valid if the precision is adequate to meet the needs of the application. Many situations with multiple significant phenomena having known laws are resolved through computer simulation. The existence of situations where the uncontrolled variables are too large is a sign that the science is incomplete. We need ways to take the other contributing factors into account.
Explanations are the way, based on our experience, we make sense of natural phenomena. They are necessary and useful, but they are models or analogies, not truths. Without a testable explanation, human creativity–making the stories of its development and its success very much worth telling a theory is unscientific conjecture.
Different explanations apply at various levels of complexity. No single explanation applies at all levels. The deeper we go into the phenomenon, eventually to the fundamental nature of matter and energy, the more elusive analogies to things we understand or experience become.
Multiple explanations of a phenomenon can be simultaneously useful. Because explanations are useful analogies and metaphors, there can be more than one way to look at a phenomenon. Each can be more or less useful depending on the circumstances. There is no need to choose.
The hallmark of scientific knowledge is that it is empirical. Laws are generalizations of observations of phenomena. To be scientific, the explanations we conceive for the phenomena must be at least theoretically testable (by observation).
The consequences of the above understanding of scientific knowledge include:
Conclusions based on laws that are applied within their tested limits are trustworthy bases for policy decisions.
The concern over the ‘truth’ of theories (which in this context means explanations) is resolved by realizing that an analogy is not the reality. Explanations help us picture the process and relate it to other things we ‘understand’. Their value is not in their ‘truth’ but in their usefulness.
Holding explanations more lightly and accepting that multiple explanations might be useful could promote imagining alternative rationalizations of the data. It could also dissuade scientists and writers from assuming an explanation is settled science before it has been empirically confirmed or hanging on to a favored explanation in the face of contrary observations.
The problem of distinguishing science from pseudo-science or non-science is resolved by science’s requirement of empirical confirmation—the data from which the laws are formulated. Hypotheses that are not at least potentially testable are not scientific.
Consequent to the arguments of this essay, there might be wide agreement that the laws proven to work within given limits are certainties within those limits. And while the explanations for the laws are likely to change, the prevailing ones are essential to understanding and advancing our knowledge of natural processes. You have seen how science, useful as it is, and as formalized as it can be, is the product of is the product of human creativity–making the stories of its development and its success very much worth telling.
A name for this philosophical position
It is worthwhile to put the above view of scientific knowledge into the framework of past and current epistemology. Thinkers as far back as Poincaré have made a distinction between laws and explanations. But they are not always treated independently in discussions of reality, particularly as the word ‘theory’ is used for either the combination or just the explanation.
My position aligns with realism. The realist Mario Bunge has coined the term scientific realism which, as he puts it, assumes the independent and prior existence of nature, shuns fictions and theories without empirical support and measurements without theories and indicators[1]. In other words, nature is real and does what it does, our formalization of its behavior is based on observation, and we must have a rational (and testable) explanation.
To this set of declarations, I add the caveat that we are only sure that our laws represent reality when they are applied within previously verified conditions. So I suggest confirmed scientific realism as a name for this view. The laws we can trust have been empirically confirmed within their tested limits, this confirmation then supported by plausible explanations.
Please spread the word that there are some things we know for sure, and that we can logically show why that is so. More than ever, we need this confirmation, and we need more scientists, teachers, and journalists to understand it.
I would like to hear from you, your reaction to these ideas and how they might apply to your life and work. Substack is built for sharing.
[1] Bunge, Mario, Between Two Worlds; Memoires of a Philosopher-Scientist, Springer International, 2016; Doing Science in the Light of Philosophy, World Scientific, 2017.