John D. Norton
Department of History and Philosophy of Science
University of Pittsburgh
https://sites.pitt.edu/~jdnorton/jdnorton.html
July 22, 2026
The content of this page is adapted from Applications of small-e empiricism, Chapter 2; and from small-e empiricism, Afterword.
This page presents a brief synopsis of the main ideas of small-e empiricism, as elaborated in my two volumes, small-e empiricism (link) and Applications of small-e empiricism (link). Both volumes elaborate considerably on the very brief sketch presented here. They include many illustration of the ideas propossed from the history of science.
Small-e empiricism preserves the defining conception of empiricism. It is that experience provides our only means of learning the truth of contingent propositions, such as form the claims of our sciences.
Big-E Empiricism, as I call it, adds to the defining conception of empiricism that the content of experience itself is all that we learn from experience. They are the only facts to which we have secure access. We cannot attribute a factual status to the results of a science that go beyond experience. A slightly softer version holds that all we learn from experience is the content of that experience itself and the possibility of factual content in further experiences of a similar character. The role in science of results beyond direct experience is purely instrumental. They serve, for example, to provide convenient summaries of our experiences and to mediate inferences from our experiences to further possible experiences.
Small-e empiricism discards this inductive skepticism in favor of the optimistic view that we can learn from experience contingent facts that are themselves remote from experience.
Traditional empiricism, as it emerged in the seventeenth and eighteenth centuries, characterized experience as the excitation of human sense organs: vision, hearing, smell, taste and touch. The subsequent processing of that experience was most commonly understood as undertaken by automatic mental processing.
Small-e empiricism allows for experiences that manifest in the excitations of human sense organs. It extends the concept to include purely instrumental sensings, such as are now commonly employed throughout the sciences. The habits of mind of a cognizer are an unreliable means of discerning the import of experience. There is no guarantee that they preserve or extract truths from experience. These habits are replaced in small-e empiricism by logical inferences, as defined by an applicable logic. Deductive inferences are those that preserve the truth of their premises. Inductive inferences are those that can preserve its likelihood.
These revisions to empiricism require three modifications to our conception of experience.
First, traditional empiricism undertook investigations that are now better treated by branches of newer sciences. The precise operation of our human sense organs is better treated by empirical physiology. The details of the habits of the mind and it processing of ideas are both better left to empirical psychology. An empiricism adapted to the practices of modern science should engage with experience as it manifests within scientific work. That manifestation is in the form of propositions in scientific writing that are declared to be reports of experience. These propositional reports of experience have been formulated so as to shed idiosyncrasies of individual human sensings and aberrations of human mental processing. They are presented as universal results that should be replicable by any competent observer.
Second, instrumental sensings expand the repertoire of what counts as experience beyond the excitations of human sense organs. Here small-e empiricism follows the near universal trend over the past century and more in the sciences of replace human sensings by instrumental sensings. Astronomers, for example, have long ceased to make primary observations by gazing into the eyepiece of an optical telescope. For over a century, their gaze and sketches of what they saw have been replaced by photographic captures of the telescopes’ findings. Such recordings provide an enduring record. Human sense organs play no role in the processes that create them.
Similarily, it is now increasingly common for the processing of these instrumentally sensed results to be conducted automatically by algorithmic computations, so that human mental process are further removed. Such algorithmic processing has become a necessity in many modern sciences. The massive volume of instrumentally collected sensings, especially in astronomy and physics, preclude human processing.
Third, experience is not characterized as a localized sensing event. Experience is conceived as a physical process that connects us with the system of interest. The processes considered are familiar. The propagation of electromagnetic waves includes propagations in the entire electromagnetic spectrum, from radio waves and microwaves, through the visible spectrum, to X-rays and gamma rays. Other propagations include sound propagation, such as the propagating noise of the 1883 Krakatoa volcanic explosion, and the propagation of seismic waves in the Earth after seismic events. These processes also include the propagation of substances. The simplest case is just our sense of smell identifying a distant, odor-producing source. Other examples include diffusion of samples in chromatographs and the material propagation of particles in cosmic rays from distant astronomical creation processes.
The various stages of these experiential processes admit propositional descriptions. The direction of physical propagation is from the system of interest towards us. Inductive inferences proceed in the opposite direction. Propositions describing stages of the process closer to us provide inductive support for propositions describing stages more remote from us, closer to the system of interest.
The reconception of experience as a process resolves one of the most vexing problems facing Big-E Empiricism. That view is founded on the idea that we can distinguish propositions representing experience from so-called “theoretical” propositions that pertain to conditions more remote from experience. Since the distinctively experiential propositions would be associated with the terminus of small-e empiricism’s experiential processes, the enduring but long-failing efforts to provide a precise characterization of them is what I call the “terminal obsession.”
The terminal obsession depends on the assumption that being designated as experiential is a monadic property of a proposition. Either it is or is not experiential. Once we reconceive experience as a process, this monadic property is replaced by a binary relation. Stages of the experiential process and propositions describing them are closer to or farther from the system of interest. There is no absolute division between the experiential and the non-experiential. Rather, propositions representing experience in scientific writing describe a conveniently chosen stage of the experiential process that is suitably close to us.
The stage reported is provisional. Doubt may be cast onto the propositional report through suspicions over the inferences to it from stages of the experiential process closer to us. It may be, for example, that these prior inferences introduced theoretical assumptions whose security is suspect. The straightforward remedy is to revert or “wind back” to a stage of the experiential process closer to us. This procedure of winding back is used routinely in science. It enables a practical escape from the exaggerated perils of what is called, in the philosophy of science literature, the “theory ladenness of observation.”
The formulation of small-e empiricism presumes this revised conception of experience and is:
Science is empirical. The propositions of experience provide inductive support for the truth of the contingent propositions of science; and the inductive support of experience is the only means of providing support for these contingent propositions.
This formulation identifies inductive inference as the mode of inference from experience. In so far as the contingent propositions of a science have a scope greater than the confines if their supporting experience, these propositions are logically stronger than the experiences that support them. It follows that the support cannot be deductive, but must be inductive.
In principle, any logic of inductive inference can be employed. However, the material theory of induction is uniquely well adapted to small-e empiricism.
The doctrine has two parts. The first asserts a positive: the capacity of inductive inference from experience to provide support for the truths of contingent propositions in science that are themselves remote from experience. What this part does not assert is that any particular science has in fact secured strong support from experience for its central claims. All the part asserts is that inductive inferences from experience have the capacity to provide such support. Whether it has been realized in any particular science must be established on a case-by-case basis.
The second part of the doctrine asserts a negative: these contingent proposition can only secure support from experience. This negative applies on the largest scale. We may have results remote from experience that are deduced from other results also remote from experience. For example, Kepler’s laws of planetary motion are deduced from Newton’s system of the world. Small-e empiricism asserts that a fuller examination of the support for these propositions collectively will find that support to reside ultimately in experience.
The immediate and most obvious argument for the positive claim is the extensive record of success of empiricism in the sciences of past millennia. Time and again experience has provided us the foundation for novel science. Empirical evidence has great power. It can force us to accept results that may initially discomfit us greatly. This was the case in the early twentieth century, when empirical results, such as concerning atomic spectra, forced an acceptance of quantum theory.
That experience has this power follows from the reconception of experience as a physical process that connects us continuously with the system of interest. For we learn of the system in the most direct way, by connecting with it. The authority of experience resides in that connection.
Finally, since small-e empiricism is inductively optimistic, it allows for strong support for the real existence of the processes comprising experience. That is, empirical investigations can affirm that there really are light and sound waves that inform us of distant systems of interest. In short, small-e empiricism supports and is supported by an ontology of experiential processes.
Big-E Empiricism is, from the perspective of small-e empiricism, an extreme, unsustainable doctrine. In additions to the difficulties indicated above, we can see that the principal arguments for small-e empiricism do not extend well to Big-E Empiricism. Support for the historical success of empirical investigations arises for both Big-E and small-e empiricism. However, because of its inductive skepticism, Big–E Empiricism cannot locate the authority of experience in the directness of the physical connection to the system of interest. For Big-E Empiricism treats these empirical processes as purely instrumental inventions. It similarly reserves judgment on whether the systems of interest are real, as opposed to another part of an instrumental scheme for relating experiences to possible experiences.
Big-E Empiricism asks us to accept the epistemic privilege of some states of excitation of human sense organs. It is true that these excitations have been fertile historically in allowing us to infer to other possible experiences. What remains unanswered why these purely mental events have this power. It cannot be explained by asserting, as does small-e empiricism, that processes like light and sound really inform us of facts in the world well beyond experience. For Big-E Empiricism is distinctive in refusing to accept their factual character. Big-E Empiricism cannot provide a foundation for the authority of experience comparable to that offered by small-e empiricism.
Arguing for the negative claim, that experience alone can provide support for the contingent results of experience, presents a greater challenge. It is not just that we must address other known modes of investigation. We must also address unknown modes. How can we preclude the efficacy of some other mode that is undiscovered and may even elude the imaginations of our greatest writers of fiction?
We can, it turns out, have some confidence that no such novel mode is nearby. Over two millennia of the most energetic investigations into nature have failed to provide a serviceable alternative; and the reason for this failure has general scope.
To begin, other modes have been considered and used. They can be categorized as Platonic insight; innate or intuitive ideas; and oracular revelation. Their universal failure has been recounted in Chapter 12 of small-e empiricism with many illustrations. That they fail is not immediately apparent from a record of their history. All these other modes claim successes. A closer examination of them, from Plato to Einstein, shows that these successes are illusory. These other modes can recover results retrospectively, once the results have already been found by empirical methods. There is however a universal failure prospectively of these modes to introduce novel results independently that survive empirical scrutiny. They are only ever wise after the fact.
A stark illustration of the comparative weakness of these other modes is given when, as is often the case, empirical investigations conflict with one of their certainties. Such conflict is inevitably resolved in favor of experience. Empirical results remain the ultimate authority.
Both empirical investigations and these other modes are fallible and can be corrected. However, the methods of empirical investigations are distinctive in containing within themselves the means to effect the correction. Empirical investigations of the experiential processes can correct mistaken appraisals of details of these processes. The more dramatic reversals of empirical science come about when new evidence arises from experience in novel domains. The reversals proceed, as empiricism requires, by accepting to the import of the evidence of experience. These other modes have no comparable means to correct their errors. They must just say that Platonic vision was clouded or innate of intuitive thoughts were somehow misapprehended.
That there are such differences between empirical and other methods can be accounted for by small-e empiricism itself. Empirical investigations provide us a rich ontology of the processes that can inform us of the world. Those that can so inform us have been exploited in small-e empiricism. The ontology is notable for lack of any processes that would allow for the successful operation of Platonic vision, or of oracles, or of innate or intuitive ideas, unless these last have in some way already been informed indirectly by experience.
What of an as yet known mode? How can it inform us of a system of interest if the mode does draw on a physical connection with that system; or draw on such a connection with other systems from which we can infer to the condition of the system of interest? Any mode that proceeds this way is, by definition, within the scope of small-e empiricism.
My hope is that readers who share my comfort with empiricism will find the version of empiricism articulated here to be a rather benign elaboration of what a modern empiricism should be. It has kept the most important part of empiricism, that we only learn contingent truths through experience, and has discarded the tendentious skepticism of stricter forms of empiricism. It has kept pace with modern science in employing an expanded notion of experience that conforms with the widespread use of instrumental sensings in science.
Where this version of empiricism may seem untroublesome and uncontroversial, the same is not true of its consequences. Taking small-e empiricism seriously leads to a cascade of corrections to present philosophy of science and epistemology. Here is a brief preview of some of them.
Inductive inferences authorize us to results logically stronger than the premises in experience from which we start. Small-e empiricism requires that this can only arise if this mode of learning is itself secured through experience. The material theory of induction conforms with this requirement, since, according to it, inductive inferences are warranted by facts that are themselves ultimately empirically founded. The same is not true of accounts of inductive inference that rely on universally applicable formal rules, in so far as those rules are not in turn derived from experience.
Since inductive inference is ineliminably fallible, we can never be assured absolutely of results learned from experience. The best we can have is very strong inductive support for our contingent results. There is no independent indicator of truth. It follows that to require truth as a condition of knowledge makes possession of knowledge opaque. Once we discard a separate requirement of truth from our concept of knowledge, the notorious Gettier cases cease to be troublesome. Gettier-like cases turn out frequently to be celebrated as scientific successes in the history of science.
Empiricism has been taken, in recent philosophy of science, to contradict scientific realism. That is only so if we adopt an inductively skeptical, Big-E Empiricism. Since small-e empiricism is inductively optimistic, it conforms with realist sensibilities. It affirms that we can learn core results in science that are remote from experience. Small-e empiricism is itself a version of realism that mediates two untenable extremes. It retracts from the temptation of strict realists to strong theses that science has somewhere, somehow overridden inductive fallibility and found final truths. And it retracts from the inductive skepticism of antirealists who base their skepticism on an exaggeration of the import of inductive fallibilism.
Finally, small-e empiricism maintains empiricism’s traditional opposition to metaphysics, in so far as metaphysics seeks to provide us with factual knowledge by means other than the empirical. For small-e empiricism maintains that no such other means work. It follows that a non-empirical metaphysics lacks sufficiently strong ways to settle debates on its major claims. The fortunes of these claims rise and fall with the rhetorical prowess of their proponents. The debates are interminable. The contrast with science is stark. Since science proceeds empirically, it has been able to secure an impressive and ever-growing body of fact.
Copyright: John D. Norton, July 22, 2026.