Skip to main content

Thomas Kuhn. The Scientific Method; strengths and weaknesses

Kuhn's description of science divides it into three basic types: immature, normal, and revoluionary. Immature science only happens once when something is new. Normal science is the most prominent aspect which consists primarily of puzzle-solving within a given paradigm. Revoluionary science is when there is instability in the scientific community and a new paradigm is called for to replace the existent one which has too many unsolved anomalies. Kuhn's description does disagree with the generally percieved view of science as truth-seeking yet he does account for the disagreements between scientists and allows that this is not necessarily irrational as it would seem to some. Yet, for many Kuhn's claims do not hold up as they are thought to be too far removed from the rationality expected of science and scientific progress.

Scientific revolutions are those times when a previous paradigm is replaced by a new, and to Kuhn conflicting, one. In debates between different paradigms Kuhn claims that the arguments are circular as 'each group uses its own paradigm to argue in that paradigms defence' (Curd, Cover 1998; 88). Even observations fail to offer a rational basis for selecting a theory as all observations rely on the paradigm from which it is being viewed. Even the meanings of the same words differ within different theories so it is necessarily the case that logical comparisons between theories is impossible as they talk past one another and so are incommensurable. As Kuhn puts it 'proponents of different theories native speakers of different languages' though it is theoretically possible to translate the theories to a degree (Curd, Cover 1998; 117). So it would seem the there is no theory-neutral means of comparison between paradigms as those advocating rival theories lack a common language. This would make the choice between paradigms clearly an issue beyond that of rational logic; if circular without shared premises then how can one paradigm be rationally and objectively better than another? Instead of rationality, it seems to be the relatively arbitrary choice of the scientific community which is at play in choosing between rival paradigms; ‘in paradigm choice – there is no standard higher than the assent of the relevant community’ (Curd, Cover 1998; 88). The choice between paradigms is more a matter of persuasion of scientists and the community than any rational argument. The scientific method of development through paradigm shifts calls for a process of decisions making which allows for equally rational men to disagree. Kuhn does though claim that, despite the fundamental conflict between competing theories, there are always a set of values which, though essentially fixed, do evolve over time with new experiences. These shared values in science consists of (but are not exclusive to); accuracy, consistency, scope, simplicity, and fruitfulness. These values though do not play the same role as 'rules'. The values can be interpreted and ranked differently by each forthcoming paradigm while rules must be followed strictly and would surely always favour one paradigm over another without great rational conflict or debate. The idea values and norms are better able to conflict while still remaining of valuable use to the scientific community. Values are also better able to provide a means of guidance in cases of conflict and are able to act as an influencial force rather than determining choice. Kuhn acknowledges that the criteria are imprecise and that they can and do conflict with one another at times, yet taking any alone is never sufficient for theory choice. Viewing the criteria for choosing between scientific theories in this way allows for Kuhn’s idea to account for the seemingly irrational aspects of science and the scientific community and also allows for there to be a shared criterion within science which can be applied even in the early stages of theory choice. These epistemic values are entirely a matter of preference and so even the shared standards of science cannot be treated as a basis for settling disputes between paradigms. Kuhn claims that every scientists choice is a mixture of objective and subjective factors yet this does not mean that scientists choices are entirely lacking in rationality. Kuhn’s view of scientific revolutions though does seem to exclude the generally perceived view of science as a practice which gathers more and more information which each time reflects closer the reality of the world. Scientific revolutions to Kuhn are ‘those non-developmental episodes in which an older paradigm is replaced in whole or in part by an incompatible new one’ (Curd, Cover 1998; 86). The theories must be accepted or rejected as a whole, one cannot merely accept some elements as paradigms are mutually exclusive. Kuhn's idea seems to portray science as though it is not actually getting anywhere closer to any 'truth' about reality but is rather merely adopting new ideas each time there is a revolution. Kuhn even goes on to claim that each new scientific revolution does not necessarily explain all the anomalies that the previous one did as the importance of each solved anomaly changes with the adoption of a new paradigm as each paradigm contains within it criteria which lays out preferences in terms of value choice. Scientific revolutions occur when the scientific community is in crisis as the accumulation of anomalies for the current paradigm is too great and so a revoluionary period is necessary in order suppress the debate to normal. These periods of scientific revoluion are relatively unstructured and sudden according to Kuhn.

Normal Scientific activity to Kuhn can be defined by puzzle-solving activities. Puzzle-solving is the everyday science which aims to solve the problems of the given scientific paradigm. For Kuhn normal science actively attempts to remove anomolies in order to support the given paradigm, scientists do not attempt to refute the paradigm with the discovery and emphasis of these anomalies as generally it is assumed by the scientific community that the prevailing paradigm is essentially correct. The paradigm is required to solve certain puzzles which would prove problematic to its reliability if unsolvable and it is Kuhn’s idea that the scientific community’s opinion is that, with the right tools and intelligence, they can be solved by the prevailing, accepted paradigm. Any failure to solve the puzzles is not blamed on the accepted theory but on the scientist who failed to solve the puzzle as it is seen as a flaw in their method rather than the theory itself. This puzzle solving though is, for Kuhn, the cumulative aspect of science as the number of puzzles solved within each paradigm increases and the paradigm seems to represent reality to a greater degree with each new ‘solution’. It is through puzzle-solving that science can be seen to improve and gain practical scientific beliefs which can be usefully applied to the world. The idea of puzzle-solving though requires that scientists select the problems that they wish to solve and they do so usually on the basis that the problem can be solved with techniques close to those which already exist. Scientists already know what they want to achieve and what result they wish to find before they begin and so ‘he designs his instruments and directs his thoughts accordingly’ (Curd, Cover 1998; 90). Kuhn’s idea of normal science portrays the scientific community as dogmatic in its teachings as it assumes the prevailing theory is pretty accurate and so all new research should confirm it and conform to it. He claims that without commitment to a paradigm normal science could not be practiced as it is through the paradigm and the areas which it does not fully explain that scientists are able to find puzzles to solve. It is puzzle-solving which distinguishes science from other practices such as astrology; each make predictions yet only science engages in acts of progressive attempts to solve the existent and problematic anomalies. Paradigms for Kuhn should be assessed on the basis of their ability to solve problems so theories should be devised which are able to do this; good theories exist within paradigms which result in the accumulation of solved problems.

McMullin proposes that Kuhn's theory would be better understood in terms of degrees of revolutions from shallow to deep. Shallow revolutions are those which do not cause any fundamental changes to occur in the theory leaving much of the theory unaltered. Deep revolutions though are paradigm shifts in which there is a great change in the methodology of the paradigm debate itself and there is resulting conflict between differing sides as to the appropriate means of assessing theories. McMullin also proposes that there cannot be such a strong divide between revolutions and puzzle-solving as these should be understood as though on a scale of different levels of intractability. There are moments when it is better for there to be a spectrum as this allows for greater flexibility in terms of the features of each being able to cross a certain boundary allowing for the claim of a higher level of rationality in some cases of paradigm debates. For McMullin, by looking at the history of science we can see that the difference between scientific revolutions and normal science is a difference in terms of degree rather than actual, easily divisable kinds as Kuhn seems to claim. The idea that the shared values within science are permanent is a bold claim which McMullin and Laudan would both contest. They would claim that values do change over time with the evolution of new theories and that there is no limit to the number of changes that occur. Kuhn does accept that changes in values occur to a degree though he would claim it to be slower and more minimal than the changes which occur within the theories themselves. McMullin also claims that Kuhns ideas regard science as a cluster concept so in order for a paradigm to be valuably scientific it is necessary that it satisfies most of the epistemic values in the cluster. The values contained within the cluster could change without essentially changing the clusters identification with science and so the values need not be permanent as Kuhn claims. Lauden agrees with Kuhn that some rules, standards, and values are highly ambigious yet he argues that Kuhn fails to show that the application of these rules or standards is invariably inconclusive. Laudan claims this is a false claim and can be demonstrated as such if we were to provide a single methodological rule for scientists to apply to concrete cases without ambiguity. In terms of scientific rules it is clear to Laudan that there are at least some which are determinate enough that many theories can obvoiusly fail to satisfy them and so we need not always provide further, subjective notions to decide between them. The idea that values often contradict one another in theory choice is not inherent to their nature, for Laudan it only happens on occasion. Kuhn fails to show why it is that the family of methodological rules should be internally inconsistent. Laudan proposes that there are actually many cases in which this is not the case. Kuhns claims that proponents of rival paradigms often disagree about what problems are most important to solve may be the case, yet the fact of this sort of disagreement existing does not neccesitate that the epistemic support of these rivals be inconclusive and thus need to be resolved by non-scientific standards. For Laudan it is wholly possible that scientific preferences can be guided by principles of evidential support which are theory neutral and not the result of individual preference. In terms of Kuhns claim to circularity, it is never explained why we cannot believe two paradigms at once if they have no shared meaning. Further, in response to Kuhns claims that meanings are theory-dependent it is possible to argue that although the sense of a term may vary among paradigms, the referent it denotes cannot surely vary so vastly as to make comparisons and experiements impossible.

Lauden claims that Kuhn offers arguments in favour of local underdetermination in which evidence and methodology is not sufficient to choose the necessarily superior theory. These arguments attempt to show rules and standards do have an influence on scientists choices, that they alone are not sufficent enough to enforce the selection of one paradigm over another. Kuhn's claim is rather sociological and non-rational when he states that it is impossible to say resistence to a new paradigm can be rationally illogical. It seems that choice between paradigms can only be non-sensical at the moment in which the theory has been wholly adopted by the scientific community. Kuhns conviction that successive theories do not resemble the truth any more than previous ones is another refutable point. He does not agree that a theory can be approximately true, for him it seems as though it is either true or false and so does not exist on a degree of propable truth. Kuhns denial of progressive science in terms of paradigms is a result of his belief that epistemic values are in essence subjective in character. Kuhns puts aside the necessity of any objective truth and instead favour theories which are best at solving puzzles. If this is the case then how does science remain cumulative and progressive when paradigms change? How is it that the number of solved problems seems to increase over time if there is no consistency between theories? Kuhn never truly addressed this issue, he leaves it standing as though it were merely coincidence and expereince which generated this. McMullen though stands as a realist and so aruges that we need to consider the truth as an aim in science in order to see why is it that scientists place so much importance on these virtues and why they have been so central in the history of science. Both Laudan and McMullin disagree with Kuhns claims that paradigms are holistic, they believe that the features of a paradigm can be rejected or accepted in parts. Lauden also claims that it is possible to have rational debate regarding which goals of scientific enquiry are more important. Kuhns focus is on non-epistemic importance as he does not place value on whether a problem is true, confirmed, or rationally credible but instead focuses on whether solving the problem would act to confirm the theory in question. When looking at the epistemic sense of importance it is clear that it is not necessary for rival pardigms to disagree about which are the imporant problems and even when this does happen it could still be the case that these issue could be resolved on the basis of shared standards and thus reason. Neither the history of science nor Kuhns arguments show that scientific revolutions cannot be resolved by rational argument based on evidence and shared rules. Neither Laudan or McMullen claim that reason, rules, and evidence always suffice to determine the outcome of such debates. Yet, they both object to the claim that they never do.

In conclusion, Kuhn's claims do account for disagreements within the scientific community and he allows for these disagreements to be rational. However, most of his claims seem too far removed from the conceptions of the aims of science. He does not allow for science to be a truth-seeking process and does not accept that progess is made in these terms. His claims of circularity also seem to fail as it seems illogical to assume that no experiements can be used which act to support one theory and disspell another; this is surely possible, scientists cannot differ so greatly in terms of understand the referent to deny the outcome. In general it seems the Kuhns claims are too bold to be applied to all science at all times. He also lacks evidence to support some of these bold claims. However, his conception of normal science does seem to have greater value in essence as it does seem as though generally the everyday methods of scientists do follow a pattern comparable with Kuhns claims. Both Laudan and McMullen refute strongly Kuhns claims about paradigm shifts yet not in such a way that his entire enterprise need be reduced to nothing. The refutations seem primarily to focus on his claims that this is always the case and that science never manages to function rationally. Perhaps sometimes and in certain cases rationality is not the primary force, but to say that it never is seems too presumptious. Thus, although Kuhns claims may allow for the social influence of the scientific community it does this to the detriment of scientific rationality and progress which undermines the point of science itself.


Curd, Martin.and Cover, J. A (eds.).1998. The Philosophy of Science: The Central Issues. New York: W. W. Norton.

Scroll to Continue


Related Articles