We have many different cognitive abilities, and human history runs on many paths. Not surprisingly, there are many ways to conduct scientific research. … These are distinct styles of reasoning, each of which has been developed in its own way, in its own time frame, and each of which contributes to the larger fabric of scientific imagination and action.Ian Hacking
In the 1980s the Canadian historian and philosopher of science Ian Hacking (1936–2023) outlined what he later called the “project of styles of reasoning—the “styles project” for short. His main objective was to investigate the historical and philosophical significance of the thesis that one can recognize the emergence and the development of a few distinct scientific styles of reasoning that emerged and stabilized at different points in time in the history of knowledge. These styles of reasoning, he averred, must be understood as different scientific rationalities, broad frameworks that govern a certain way of investigating the world, and involve new types of evidence, methods of inquiry, and questions. Although Hacking has added several constructive suggestions to his original nucleus of ideas over the last forty years, his original thesis has never been fully developed into a comprehensive theory. In other words, his styles project has remained a set of suggestions scattered in different papers to be developed and subjected to critical scrutiny.
My book History of Rationalities: Ways of Thinking from Vico to Hacking and Beyond can be placed between an assessment of Hacking’s claims and an attempt to develop, correct, and present the styles project in a more systematic way to make something coherent and complete out of it. Thus, first of all, my book can be viewed as a history of rationalities from the first civilizations to the present day based on some of Hacking’s suggestions about how, in certain epochs, human beings thought and found out about our world and altered it. However, since to infer philosophical implications from this history of rationalities and analyze them has been central to my methodology, it is possible to say that my book is also a “theory of scientific rationalities,” that is, an analytical account of the styles project and its philosophical implications.
Styles of Scientific Reasoning
The history of rationalities presented in the book consists of a description of the genesis and the historical development of six styles of reasoning still practiced in modern science:
- 1) the “algorithmic style of reasoning,” which goes back to civilizations more ancient than Greece whose methods of calculation followed step-by-step lists of rules;
- 2) the “postulational style of reasoning,” which emerged in ancient Greece and is exemplified by Greek geometry; it consists of proofs by deduction on the basis of explicit axioms;
- 3) the “statistical style of reasoning,” which emerged in the seventeenth century and is characterized by the use of the statistical analysis of population and the calculus of probabilities;
- 4) the “laboratory style of reasoning,” which is characterized by the method of experiment and measurement, which took shape in the Scientific Revolution;
- 5) the “historico-genetic style of reasoning,” exemplified by Darwin’s argument in The Origin of Species; and
- 6) the “taxonomic style of reasoning,” that is, the ordering of variety by comparison and taxonomy, which reached its acme in the era of the botanist Carl Linnaeus (1707–1778).
Hacking mentioned all these styles in different writings, but he has only given an approximate description of two of them: the statistical and laboratory styles. For my part, I have shown that the six styles above share some features that Hacking attributed either to the statistical or laboratory style of reasoning:
- a) being ways of thinking and doing, in particular ways of intervening in the world, that is, not only ways of finding out “that” but also ways of finding out “how to”;
- b) relying on their own standard of evidence for finding out about the world;
- c) introducing new sentences that are candidates for truth-or-falsehood as well as new types of explanations and/or new criteria, laws, and classifications;
- d) being self-authenticating, that is, for certain results achieved by using a given style there is no other standard of correctness than that style;
- e) representing a sharp break in the history of Western thought.
Towards a Theory of Scientific Rationalities
To better understand what these styles of reasoning are, I will provide just a few hints of the account in the book. Before the late fifth century, the conditions for the emergence of the deductive style of reasoning had not been met. The competition within the Greek polis created the need for a form of argument that would go beyond mere persuasion. When the proof emerged, it came to be considered the paradigm of what it meant to settle an argument. Thinking in the deductive style meant following a two-step process: to search for the simplest and fewest premises and to deduce their implications by “proof.” The deductive style of reasoning introduced new sentences that became candidates for truth-or-falsehood, that is, the very sentences used to express the geometrical a priori propositions, which could not have that sense unless they were embedded in the practice of geometrical demonstration itself.
While the ancient Greeks pursued scientific knowledge systematically and for its own sake, the Babylonians and the Egyptians acquired and transmitted only sub-scientific knowledge, that is, specialist knowledge pursued only in view of its applicability. For example, Babylonian algebra consisted of step-by-step lists of rules-algorithms for representing algebraic equations and, ultimately, for computing them. To adopt the algorithmic way of thinking means following methods of calculation-a step-by-step list of rules that might be represented by a formula.
In The Emergence of Probability, Hacking argued that the concept of probability was absent before the time of Blaise Pascal (1623–1662). What made the concept of probability possible was the appearance of a new form of evidence. I have used these important points to outline the genesis and development of the statistical style of reasoning. In medieval thought, scientia represented the knowledge of universal truths obtained by demonstration; conversely, opinio was associated with beliefs, resulting from argument or disputation, which could not be demonstrated. The word probabilis did not suggest the idea that a hypothesis was better supported by evidence than others but meant “worthy of approbation.” Later on, in the writings of Renaissance physicians, opinions were probable when supported by signs of Nature rather than the written word. Besides approbation, there was no other concept of evidence related to opinion. What was lacking was the evidence provided by things, not to be confused with the evidence provided by the data of the senses. To use the example given by J. L. Austin (1911–1960) and quoted by Hacking, pig-like marks and buckets of food outside a sty represent the evidence of things for the statement that there is a pig in the sty. For us, books and testimony represent indirect evidence reported by other people. The Renaissance had the order reversed: testimony and authority were primary; things counted as evidence only insofar as they resembled the authority of testimony and books. The physician Girolamo Fracastoro (1483–1553) wrote that, among the signs in the sky, air, soil, and water are premonitory. These changes opened the way for the evidence of signs to turn into the evidence of things. Since not all signs are to be trusted with certainty, the idea of probability is connected with frequency, that is, with what “almost always” happens. Here, we already recognize some of the features of our statistical concept of probability. In a text published in 1650 by Thomas Hobbes, the concept of evidence of things conjoined with that of frequency had already taken full shape. In 1662, the Port Royal logic as well, published by members of the Jansenist movement, distinguished arbitrary and conventional signs—the concept of a sign as evidence had become endemic; stable and law-like regularities became worthy of observation. Later on, the studies of the Jacques Bernoulli (1654–1705) culminated in the central limit theorem; in 1756 Thomas Simpson (1710–1761) applied the theory of errors to the discussion of uncertainty; finally, in the 1820s the first public statistics were published.
The concept of the laboratory style of reasoning is meant to account for the emergence of a community of people that used a new type of evidence: they utilized devices to produce phenomena (effects) that did not previously exist in isolation. When did it start? The production of effects settled into routine in scientific research in the seventeenth century. In Leviathan and the Air Pump, Steven Shapin and Simon Schaffer explain that in the late 1650s, the natural philosopher Robert Boyle (1627–1691), one of the founders of the Royal Society, instructed Robert Hooke (1635–1703) to build him an air pump: a closed vessel (receiver) from which air could be extracted by using a pump mechanism. Boyle produced an approximate vacuum in the receiver to be able to perform a series of experiments to investigate the composition and compressibility of air. These experiments were carried out in a new place, the laboratory, namely, a site for finding out by making something: for example, finding out that air is necessary for a candle flame to continue burning by “creating” the phenomenon of the vacuum in a vessel. The nascent laboratory was a “public” space: being public meant that individuals attended the performance of an experiment; they validated and constituted a piece of knowledge by agreeing about what they saw. Conversely, the alchemists produced their knowledge claims in a private and undisciplined space.
The historico-genetic style of thinking can be summarized in these steps: 1) assuming that things stand in a certain way in some domain of interest; 2) devising an explanation for how things got to be that way; 3) inferring from the power of the explanation and its superiority to any other available candidate that it tells how things are. “Darwin’s long argument,” as Darwin called his own argument in The Origin of Species, can be taken as an example of the historico-genetic style of thinking. Indeed, the explanation of the adaptation of living beings is considered the primary fact to be explained by the theory of evolution (step 1). William Paley (1743–1805) had proposed a different explanation according to which adaptation must have been “designed” by God. As there are several facts that are not explained by Paley’s argument but are explained by the “long argument,” one can conclude that Darwin proposed the best explanation (step 3).
Finally, the way of thinking of taxonomists from Aristotle to modern taxonomists is characterized by three mental actions: grouping, ranking, and systematizing. Grouping and ranking require criteria such as those based on similarities in Aristotle’s and Linnaeus’s classifications. Systematizing consists of a causal explanation to interpret the classification: for example, the fixity of species for the great chain of being or the evolutionary theory for the Linnean classification. Linnaeus represents the point in time in which taxonomic thinking suddenly came to dominate the scientific milieu. Taxonomists themselves consider their methods as ways of finding out. This is evident not only in Aristotle, for example, when he wanted to explain the nature of kinds by identifying their place in nature, but also in the thought of later students of classification. In the mid-seventeenth century, the idée fixe of taxonomists was “to name means to know”: for Linnaeus, denominatio was one of the aims of botany; classifying constituted a way of finding out the structure of divine thought.
Philosophical Implications
As I have mentioned, my book History of Rationalities draws a number of philosophical implications from this characterization of the notion of style and analyzes them. How did certain concepts emerge and change? How have our styles of reasoning changed and why? To what extent is the emergence of the styles of reasoning at a certain point of history a contingent circumstance? Why have these styles of reasoning endured? Readers can find in my book an attempt to answer these questions. To give an example, a question I address is what the notion of style implies about the future evolution of our knowledge. I argue that the knowledge we acquire by adopting different styles has no unity and cannot converge on any kind of unitary description of the world.
It is for this reason that Hacking’s metaphor is appropriate: there are “heapings up” of knowledge because each style accumulates new laws, phenomena, facts, and objects that cannot be connected by a single description of the world. This conclusion reinforces and expands the premise of the history of knowledge as currently practiced by historians, that is, the contention that there are different kinds of knowledge generated by different actors in different historical settings, and none of these knowledges is superior to another. My book adds that the kind of knowledge pursued by scientists is obtained through a plurality of different styles of reasoning, each of which has its own standard of truth. There is no other higher standard to which styles directly answer. Furthermore, the idea that science will converge on a single true, coherent, and complete description in the long run is at odds with the mechanism by which the styles of reasoning produce knowledge: the growth of many different epistemological items is open-ended and never final and complete.
It cannot be forgotten that Hacking’s notion of various styles of reasoning is only one of the many models of the history of knowledge that thinkers have proposed at different times. For example, in The Order of Things, the French philosopher Michel Foucault (1926–1984), with his notion of “episteme,” essentially highlighted the structures common to the knowledge practices of a historical period. For his part, the American historian and philosopher of science Thomas Kuhn (1922–1996), in The Structure of Scientific Revolutions, put forward the notion of a “paradigm” as a general framework underlying scientific endeavors with components such as exemplars, ontological assumptions, and values. In my book, it is also possible to find a comparative analysis of the most important notions that, like those of Hacking, Foucault and Kuhn, imply the existence of different ways of thinking. A question I have addressed is whether in some cases a given notion provides the most enlightened ordering of the historical record or a more profound understanding of the reasons concepts or scientific discoveries become objective.
Luca Sciortino teaches Philosophy of the Human Sciences at the university UniTreEdu in Milan, Italy. His research concerns the history and philosophy of scientific thought and especially the field of “historical epistemology”. He earned his PhD in Philosophy of Science from the Open University (UK) and has been a postdoctoral researcher at the University of Leeds, Centre for History and Philosophy of Science (UK). Besides the book presented here, he has authored numerous research articles published in international peer-reviewed journals such as Erkenntnis, Studies in History and Philosophy of Science, and the International Journal in Philosophy of Science.
