Determinism vs. Underdeterminism in Scientific Change
Is the process of scientific change a strictly deterministic process? Will two unconnected communities experience a similar historical series of changes in their individual mosaics?
The process of scientific change is not restricted to any single community. As such, the evolution of different mosaics could take two very different paths. This question seeks to uncover whether or not separate communities would experience similar stages of change in their mosaics and to elucidate under which circumstances such changes can occur.
In the scientonomic context, this question was first formulated by Hakob Barseghyan in 2015. The question is currently accepted as a legitimate topic for discussion by Scientonomy community.
In Scientonomy, the accepted answers to the question can be summarized as follows:
- Transitions from one state of the mosaic to another are not necessarily deterministic. Scientific change is not a strictly deterministic process. The process of method change is not necessarily deterministic: employed methods are by no means the only possible implementations of abstract requirements. The process of theory change is not necessarily deterministic: there may be cases when both a theory's acceptance and its unacceptance are equally possible.
There are, broadly speaking, two stances concerning the determinism of scientific change. One is that scientific change is purely deterministic – i.e. that given a scientific mosaic at some time, it will evolve in one and only one fashion. This thesis states that “two unconnected scientific communities will necessarily pass through similar stages” 1, and that there is no room for sociological or other extra-scientific factors to influence the evolution of the accepted theories and employed methods of a given community. This position has been held by many philosophers since at least Leibniz, as pointed out by Laudan – although it is likely that this view would be accepted by much earlier figures such as Plato or Aristotle.2 1 According to Laudan, most philosophers subscribed to scientific determinism (what he calls the “Leibnizian ideal”) only up until recently. Early and mid 19th century philosophers like Mill, Herschel and Whewell would subscribe to determinism, with Mill believing not only that scientific change is deterministic, but that we actually have the explanatory rules.2 Likewise, logical positivists and the likes of Carnap, Reichenbach, and Popper believed in scientific determinism, and strove to find the rules that actually determine scientific change. 2
This traditional position is opposed by the view that scientific change is not strictly deterministic – i.e. that given a scientific mosaic at some time, it can evolve in many ways, and there are many ways it could have come to be. Although this has not been the mainstream position in the past, not all prominent thinkers subscribed to the deterministic view – Descartes challenged it (albeit not specifically related to scientific change) with his evil demon, and Mill actually expressed a lucid early description of what would become the full-fledged thesis of scientific underdeterminism: “an hypothesis…is not to be received as probably true because it accounts for all the known phenomena, since this is a condition sometimes fulfilled tolerably well by two conflicting hypotheses.”34 Scientific underdeterminism is supported by the SSK movement, Laudan, Duhem and Quine, the latter two taking a more radical stance.4 While determinism is relatively easy to describe, underdeterminism is more subtle, and there a number of different stances one may take while still being said to support scientific underdeterminism. In order to sort this out, it is instructive to make the traditional distinction between so-called “holist underdeterminism” and “contrastive underdeterminism”.4
Holist underdeterminism is said to occur when we are unable to test a specific hypothesis in isolation. For example, when testing Einstein’s general relativity, astronomers attempted to observe light bending during a solar eclipse – they did in fact observe this light bending, and used this as evidence to confirm Einstein’s theory and to reject Newton’s. But, strictly speaking, the evidence given by this experiment underdetermines our response to it, because it could be that this test refutes theories of optics rather than Newton’s theory of gravity, or that it refutes our estimates of the mass of the sun, or any other connected, tacitly assumed theory. The key point behind holist underdeterminism is that there can be no such thing as a “crucial experiment” – an experiment that decisively confirms one theory and refutes another. This idea was developed by Duhem in his Aim and Structure of Physical Theory, and was further developed by Quine in subsequent years.4
Contrastive underdeterminism is more relevant to our discussion on the thesis of scientific underdeterminism. It states that “for any body of evidence confirming a theory, there might well be other theories that are also well confirmed by that same body of evidence”4 – where the meaning of “confirmed” is of course dependent on the given scientific community. It may be better to state this as “for any body of evidence causing a theory to satisfy a communities acceptance criteria, there might well be other theories…” The reason this form of underdeterminism is more relevant to the issue at hand is because it clearly states that there are many possible theories that could satisfy a given body of evidence – and that as a result, there may be no way of predicting which of them is accepted. A similar thesis can be formulated for methods in the obvious way – given a community’s implicit expectations, there are many methods that can implement these, so there is no way to predict which of them will become implemented.
|Community||Accepted From||Acceptance Indicators||Still Accepted||Accepted Until||Rejection Indicators|
|Scientonomy||1 January 2016||That is when the community accepted its first answer to this question, the Scientific Underdeterminism theorem (Barseghyan-2015), which indicates that the question is itself considered legitimate.||Yes|
|Underdetermined Method Change theorem (Barseghyan-2015)||The process of method change is not necessarily deterministic: employed methods are by no means the only possible implementations of abstract requirements.||2015|
|Underdetermined Theory Change theorem (Barseghyan-2015)||The process of theory change is not necessarily deterministic: there may be cases when both a theory's acceptance and its unacceptance are equally possible.||2015|
|Scientific Underdeterminism theorem (Barseghyan-2015)||Transitions from one state of the mosaic to another are not necessarily deterministic. Scientific change is not a strictly deterministic process.||2015|
If an answer to this question is missing, please click here to add it.
|Community||Theory||Accepted From||Accepted Until|
|Scientonomy||Scientific Underdeterminism theorem (Barseghyan-2015)||1 January 2016|
|Scientonomy||Underdetermined Theory Change theorem (Barseghyan-2015)||1 January 2016|
|Scientonomy||Underdetermined Method Change theorem (Barseghyan-2015)||1 January 2016|
In Scientonomy, the accepted answers to the question are Underdetermined Method Change theorem (Barseghyan-2015), Underdetermined Theory Change theorem (Barseghyan-2015) and Scientific Underdeterminism theorem (Barseghyan-2015).
Underdetermined Method Change theorem (Barseghyan-2015) states: "The process of method change is not necessarily deterministic: employed methods are by no means the only possible implementations of abstract requirements."
The third law allows for two distinct scenarios of method employment. A method may become employed because it follows strictly from accepted theories or employed methods, or it may the abstract requirements of some other employed method. This second scenario allows for creative ingenuity and depends on the technology of the times, therefore it may be fulfilled in many ways and allows underdeterminism 1.
Underdetermined Theory Change theorem (Barseghyan-2015) states: "The process of theory change is not necessarily deterministic: there may be cases when both a theory's acceptance and its unacceptance are equally possible."
The process of theory assessment under the TSC is underdetermined for two reasons. First, only theories that are constructed are available for assessment. Whether or not a theory is ever constructed is, at least partly a matter of creativity, and is therefore outside the scope of the TSC. Second, it is at least theoretically possible that a process of theory assessment will be inconclusive. This might be because the requirements of the method employed at the time might be vague (e.g. Aristotelian requirements of "intuition schooled by experience").1
Scientific Underdeterminism theorem (Barseghyan-2015) states: "Transitions from one state of the mosaic to another are not necessarily deterministic. Scientific change is not a strictly deterministic process."
Scientific underdetermination is the thesis that the process of scientific change is not deterministic, and science could have evolved differently than it did. Hypothetically, two scientific communities developing separately could experience an entirely different sequence of successive states of their respective scientific mosaics. Even without the TSC, the implausibility of scientific determinism can be seen by considering the process of theory construction, which is outside the present scope of the TSC. Theory construction requires creative imagination, and the formulation of a given theory is therefore not inevitable. Still, underdetermination can also be inferred as a theorem from the axioms of the TSC.1
This deductive inference occurs as follows. The theorem is the consequence of two related theses: that of underdetermined method change, and that of underdetermined theory change. Since method change and theory change are exactly the transitions in the scientific mosaic, showing that neither method change nor theory change is deterministic is sufficient to imply the Scientific Underdetermination theorem (SUT).
The underdetermination of method change follows from the 3rd Law: “A method becomes employed only when it is deducible from some subset of other employed methods and accepted theories of the time.” As a result of this law, a method can be employed in two ways: either it strictly follows from other accepted theories and employed methods, in which case the change is, in fact, deterministic, or it implements the abstract requirements of some other employed method. In the latter case, the change is underdetermined since abstract requirements can give rise to many different implementations. As a concrete example, we have an accepted theory that states, “when counting the number of living cells, the resulting value is acceptable only if it is obtained with an ‘aided’ eye.” A number of different methods can implement this abstract requirement, like the plating method or the counting chamber method.1 Thus the method is underdetermined by the abstract requirements, so the process of method change implementing these requirements is not deterministic, which is exactly the statement of the underdetermination of method change.
The underdetermination of theory change comes from the 2nd law and the possibility of inconclusive theory assessment. The 2nd law states that “In order to become accepted into the mosaic, a theory is assessed by the method actually employed at the time”. This assessment can result in conclusive acceptance, conclusive rejection, or it can be inconclusive. In both conclusive cases, the theory change is deterministic, but if the theory assessment is inconclusive, then the theory can be accepted or rejected. So the process of theory change is not necessarily deterministic, since it is, in fact, possible for assessment to be inconclusive. The question of whether there actually exist cases of inconclusive theory assessment is a task for the history of science, and is irrelevant to our discussion.
These two theses combine to form the SUT, since changes in theories and methods are all the transitions that occur in the scientific mosaic, and we have seen that the underdetermination of theory and method change follow deductively from the 2nd law, the 3rd law and the possibility of inconclusive theory assessment.
This question is a subquestion of Mechanism of Scientific Change.
- a b c d e f Barseghyan, Hakob. (2015) The Laws of Scientific Change. Springer.
- a b c Laudan, Larry. (1984) Science and Values. University of California Press.
- ^ Mill, John Stuart. (1974) The Collected Works of John Stuart Mill, Volume VII - A System of Logic Ratiocinative and Inductive, Being a Connected View of the Principles of Evidence and the Methods of Scientific Investigation (Books I-III). University of Toronto Press, Routledge and Kegan Paul. Retrieved from http://oll.libertyfund.org/titles/246.
- a b c d e Stanford, Kyle. (2016) Underdetermination of Scientific Theory. In Zalta (Ed.) (2016). Retrieved from http://plato.stanford.edu/entries/scientific-underdetermination/.