A Detailed Introduction to the Philosophy of Science

Let’s take your understanding of science to the next level

Nico Ryan

Published in

The Understanding Project

23 min readJul 30, 2019

All through high school I struggled with math — big time.

I barely scraped by with a grade of 70% each year, despite the fact that I dedicated hours of study to the subject every day and worked with a paid tutor who was himself a high school math teacher!

Numbers, equations, geometry, and the like are not my thing.

For some people, my math is their philosophy.

Some folks simply can’t stand thinking, reading, writing, or conversing about philosophical topics; it’s one of the last things they’d ever choose to do.

If you’re not a fan of philosophy and you prefer to learn about a subject from a top-down, macro perspective rather than from a bottom-up, micro perspective, detailed discussions about truth, fallacies, fact-value distinctions, and objectivity versus subjectivity might not resonate with you.

It’s possible that such discussions don’t resonate with you because they can’t resonate with you — at least not yet: if you don’t understand the words used in an article—words like dialectic, empiricism, post-truth, and relativism — there’s little sense in continuing to read the piece.

Philosophers write about a wide range of concepts, some of which are specialized and circumscribed.

If you come across a piece of philosophical writing that explores specific ideas but you’re not adequately versed in the concepts on which the ideas are based, you’re likely to find the writing boring, confusing, or intolerable.

If this is the sort of experience you have whenever you try to read philosophy, I’m here to make things easier for you.

If you want to learn more about philosophy and science, and if you often struggle to grasp the details of philosophical content, this article is for you.

In this piece, I provide an introduction to a number of the central concepts, ideas, and questions upon which the philosophy of science rests.

You’ll get much more value and enjoyment out of philosophical texts if you have a decent understanding of — or at least some familiarity with — the conceptual underpinnings and key debates upon which these texts rest.

The philosophy of science is worth examining because it is here that some of the most interesting and difficult questions about science are being explored.

(To simplify the discussion, the terms philosophy of science and metascience are used interchangeably. The meta in metascience means beyond, above, or at a higher level than science.)

Table of Contents

1. Mouton’s Three Worlds Model
2. The Meaning, Practice, and Purpose of ‘Science’
3. Ontology and Epistemology: The Two Foundations of Metascience
4. The Importance of Ontology
5. Stratified Versus ‘Flat’ Models of Reality
6. A Few Words About Methodology and the Myth of the ‘Scientific Method’
7. Summary
8. Recommended Readings

Mouton’s Three Worlds Model

The best way to begin exploring the philosophy of science is by differentiating it from several other areas of inquiry.

Mouton’s 1996 text, Understanding Social Research, offers a helpful starting point in the form of a three worlds model of human inquiry.

The model differentiates the different rules, practices, role expectations, and sets of skills and competencies associated with three realms of knowledge production.

Each of the three realms comprises 1) a unique form of knowledge and 2) a corresponding epistemic interest, i.e., a purpose, goal, or function of knowledge creation.

World 1 is that of everyday life and lay knowledge:

Peoples’ routine engagements with learning, experience, and self-reflection contribute to stocks of problem-solving knowledge directed at coping with daily tasks and challenges (p. 8).
World 1 consists of the objects of everyday life: individual human beings; groups; social practices; institutions; and the elements of our physical surroundings — in other words, the ‘stuff’ we typically consider to make up reality (p. 10).

World 2 is that of science and truth:

Here, the phenomena of World 1 (i.e., everyday life) are transformed into objects for systematic and rigorous study, i.e., into scientific objects of investigation (pp. 8–9).
The motivating epistemic interest here is the desire to produce a truthful understanding of the operations of the natural and social worlds (pp. 8–9).
World 2 comprises the academic disciplines of the social, human, and natural sciences (criminology, psychology, sociology, etc.); different methods for investigation (e.g., qualitative versus quantitative techniques); scientific hypotheses, models, and theories; and all the objects of World 1 (p. 10).

World 3 is that of the philosophy of science (metascience) and critical/reflective inquiry:

Here, the phenomena of World 2 (i.e., science) are explicitly transformed into objects of reflective inquiry driven by a critical interest, i.e., by “[the] aim to criticise, dissect, deconstruct or analyse what scientists do toward the ultimate improvement of science [itself]” (p. 9).
World 3 contains academic disciplines like the philosophy of science, the sociology of knowledge, and the history of science; diverse methodologies guiding scientific inquiry (e.g., positivism versus interpretivism versus critical realism); and all the objects of World 2 (p. 10).

To recap, Mouton insists all knowledge can be classified into 1) everyday (problem-solving) knowledge, 2) scientific (truth-based) knowledge, and 3) metascientific (critical/reflective) knowledge.

How can Jane successfully apply for a new job? → World 1 (coping with everyday problems)
How can scientists demonstrate a new medication is safe enough for human consumption? → World 2 (empirical scientific study)
How can scientists or philosophers make sense of what science and scientists as a whole do or should do? → World 3 (investigation into science as such)

With this distinction in place, let’s now look at the philosophy of science (World 3) in greater detail.

The Meaning, Practice, and Purpose of Science

The philosophy of science is concerned with the very definition of science: what science is; what scientists should try to accomplish; and how science operates both in theory and in practice.

Bourdieu (1991, p. 14) puts it like this:

“[Metascience studies] the principles of the construction of the object of study as a scientific object and the rules of delimiting the relevant problems and methods that must be employed to resolve them and to measure accurately the solutions.”

Keat and Urry (1975, p. 209), using more straightforward language, explain it using these words:

“[Metsacience explores] what sorts of belief and enquiry are to count as science: and this is a philosophical issue that cannot be avoided or eliminated by any amount of sociological, psychological, or historical investigation.”

Thus, the philosophy of science — the study of science as such — seeks to

establish which problems (puzzles, questions, etc.) scientists can legitimately treat as scientific problems
establish which forms of reasoning, concepts, and methodological instruments scientists can legitimately use to investigate such problems.

Whenever we’re talking about the results of a particular study, we’re in the realm of science.

Whenever we’re talking about the general parameters according to which the results of any scientific study (or, indeed, the theories that a study uses) can be judged as either accurate and defensible or flawed and unjustifiable, we’re in the realm of metascience.

In the course of their everyday activities, scientists don’t always think about the assumptions, concepts, ideas, instruments, and theories they use to make sense of themselves and others, of the world around them, of their study objects, and of the parameters of their investigations.

Sometimes they do, but many times they don’t: they’re too busy doing science to spend time thinking about science.

One consequence is that scientists are often unaware of the various things they take for granted as they go about their everyday scientific business (Collier, 1994, p. 17).

Scientists’ taken-for-granted concepts, ideas, and theories are a key focal point for the philosophy of science.

Many of the foundational beliefs on which scientists unconsciously rely to carry out their day-to-day work belong to two sub-areas of metascience known as ontology and epistemology.

Debates about truth, fallacies, fact-value distinctions, standards of evidence, objectivity versus subjectivity, and a whole host of other philosophical and scientific concepts and theories are, in fact, debates about ontology and epistemology.

We need to spend some time looking at the domains of ontology and epistemology because they are the realms of study from which many of the topics that philosophers explore derive.

Ontology and Epistemology: The Two Foundations of Metascience

First, here are a couple of simple descriptions:

Ontology is the study of what exists; it’s concerned with ‘stuff’.
Epistemology is the study of what (and how) we can know about what exists; it’s concerned with the production and limits of human knowledge.

Ontology is the domain of metascience that investigates existence (being) as such.

Ontological theories try to offer ways of analyzing that which is (that which exists), the nature of (different forms of) reality, and the kinds and numbers of relations that persist between entities.

Ontology asks questions like these:

What is reality made of? How is reality structured?
What sorts of things can scientists investigate?
What types of associations exist between objects?

Whenever scientists or philosophers express judgments about the nature of a thing to be explained, they’re engaged in ontological examination.

Epistemology is the realm of metascience that examines knowledge production.

Epistemological theories try to present ways of thinking about and assessing what can be known, how human beings can generate knowledge, which methods and tools can be used to develop knowledge, and what the nature and status of scientific results are.

Epistemology asks questions like these:

What is knowledge? How, if at all, is knowledge different from information and/or from understanding/comprehension?
Are there differences between efforts to produce explanations of causal mechanisms on the one hand and interpretive understanding of meaning on the other? If so, what are these differences?
How can scientists employ techniques — modes of logic and methodological instruments — to come to know something?

Whenever scientists or philosophers endeavour to outline or assess the procedures through which some object or process can become known, they’re involved in an epistemological pursuit.

Epistemology also involves the study and use of what are known as modes of inference (or modes of reasoning).

In basic terms, modes of inference are “thought operations, i.e., different ways of reasoning and thinking in order to proceed from something to something else” (Danermark et al., 2002, p. 73).

They offer different conceptual and logical procedures for transitioning between the general and the particular and vice versa, and they seek answers to questions such as, “What does this mean?” “What follows from this?” and “What must exist for this to be possible?” (Danermark et al. 2002, p. 79).

You’re probably familiar with deduction and induction. It’s worth pointing out that scientists sometimes use other, less-known modes of inference, i.e., abduction and retroduction (Danermark et al. 2002, p. 79–106).

Insofar as questions about truth and knowledge often relate to questions about the validity of different forms of reasoning, understanding the differences between, and the strengths and limitations of, each mode of reasoning is essential to grasping the details of many epistemological problems.

One last key point to recognize, which many (positivist) scientists misunderstand or under-appreciate, is that “every epistemology contains within it an ontology” (Rigakos and Frauley, 2011, p. 253).

This means that any theory on what knowledge is, how knowledge can be attained, or what the limits of human knowledge are necessarily contains within it assumptions about the nature of reality, about what ‘stuff’ can be made amenable to scientific investigation, and about the structure and other attributes of the object under study.

This is because questions about how to study x are necessarily shaped by answers to questions about what x is: “it is the nature of objects that determines their cognitive possibilities for us” (Bhaskar, 1979, p. 27)

We can’t know a triangle has three sides if, in fact, a triangle has more or fewer than three sides.

Nevertheless, scientists sometimes fail to grasp how their implicit beliefs about ontology — what exists? how are things connected to each other? what is the nature of the study object? — contradict or undermine their explicit beliefs about epistemology and/or their use of methodological tools.

If you want to learn more about why this is problematic, feel free to read up on the critical realist objections to positivism, especially regarding the epistemic fallacy and the ontic fallacy (Collier, 1994, pp. 70–104).

The Importance of Ontology

Although they’re not always framed as such, many scientific questions, debates, and puzzles are rooted in efforts to examine and make sense of the nature of reality and the objects and processes within it.

In other words, ontological problems rest at the heart of science.

On one level, this is obvious. After all, scientists are constantly trying to explain the ‘stuff’ they investigate:

Physicists aim to uncover the nature of matter.
Neuroscientists endeavour to explain how the brain works.
Sociologists try to discover the structures and untangle the influence of social relations, social interaction, and elements of culture.

On a deeper level, though, science isn’t just about designing and carrying out specialized, concrete empirical studies in an effort to ‘build up’ knowledge over time.

(See Mills (1959, pp. 65–67) for a discussion of why the building-block approach that positivist scientists favour is flawed.)

To do their day-to-day work, scientists have no choice but to embrace — consciously or unconsciously — theories about the nature of the world.

It is because they accept that reality is a certain way that scientists feel justified in asking and investigating certain questions, using certain methodological tools, offering certain interpretations of data, and defending or attacking certain scientific studies.

It’s also why philosophers (and others) who write about knowledge, truth, fallacies, and so on make the specific arguments they make.

For example, if you accept the correspondence theory of truth — which insists that the truth or falsity of a statement is determined by whether the statement accurately reflects the state of the world as it actually exists — you accept it, for one, because you believe reality is such that human beings can ‘access’ it directly enough to know how it operates.

If you don’t believe in the correspondence theory of truth, one likely reason is that you’re convinced there’s an impassable gulf between reality ‘out there’ and what human beings can know, where what human beings can know is a product of the kind of thing they are.

Many similar disputes in science and philosophy can be boiled down to the claim that some given assertion can’t (or is unlikely to) be true because ‘that’s just not how the world works’. And how the world works is, of course, a matter of ontological examination.

Within the philosophy of science, there are many interesting debates about the structure of reality and the entities and processes within it, debates that should inform mainstream scientific practice much more than they do.

Let’s look at one of these debates in detail, one that is fundamental to a number of the issues about which philosophers, social theorists and sociologists, and science writers tend to write.

Then, to end the article, let’s briefly discuss methodology and dispel the myth that the Scientific Method is the definition of ‘good science’.

Familiarizing yourself with the ideas explored within the next two sections will provide you with some foundational knowledge to better understand and engage with arguments about truth, social structure, power relations, (so-called) validity and reliability in science, and other important matters.

Stratified Versus Flat Models of Reality

One of the key questions with which ontology is concerned is, “What is the nature of reality?” Another way to express this question is to ask, “What is the ontological organization of the world?”

We can identify at least two contrasting ontological models:

A flat ontology
A depth or stratified ontology.

The two models are worth considering because they allow scientists to design and carry out studies and to interpret and analyze data with differing degrees of sophistication, nuance, and complexity.

A flat ontology

is the defining ontology of mainstream science
suggests all reality exists on one main ‘level’
treats what can be measured as equivalent to both what exists and what is scientific.

Within a flat ontology, different dimensions of reality are recognized, but, ultimately, everything is understood to exist on the same ‘plane’.

The most common way in which scientists (implicitly) advocate for a flat ontology is by insisting that what exists is synonymous with what can be measured.

If something can’t be measured, it isn’t real, and it isn’t real because it can’t be transformed into an object of scientific investigation, which produces scientific knowledge.

On this view, there’s nothing beyond the empirical world with which science is rightly concerned; ‘metaphysical’ discussions of unobservable entities or processes don’t belong in science (Keat and Urry, 1975, p. 82).

Yes, researchers investigate many aspects of reality that aren’t directly observable — such as intelligence, conscientiousness, and authority — but the use of ‘operational definitions’ allows these unobservable items to be transformed into measurable (and thus scientific) phenomena (Hoyle, Harris, and Judd, 2011, pp. 75–78).

These, then, are the foundational principles of a flat ontology, and they give rise to what is known as empiricism.

Empiricism, which is not the same thing as empirical, is the notion that “anything worth knowing…must be apprehendable to our senses” (Rigakos and Frauley, 2011, p. 243) and “sense-perception exhausts the possible objects of knowledge” (Bhaskar, 1986, p. 230).

Empiricist science is widespread today, with many scientists embracing empiricism without even realizing it.

The problem with empiricism is that acceptance of a flat ontology necessarily leads to incomplete and fractured scientific investigation and analysis:

“[T]hat which is in fact real is greater than what human beings might be able to directly experience, including processes that have a hand in the emergence of actually existing things. Criminologists, for example, frequently speak of ‘power relations’. These are not directly observable, but the effects of domination and subordination, wage inequality, preferential treatment, racism, and the like are experienced. We might suggest that these are embodied forms of relations of power.
Analyzing the emergence of inequalities rather than the experience of inequalities themselves would take us beyond the realm of the empirical[.] … This is important as the things that are of concern to…scientists are not necessarily captured, revealed, or exhausted by our experiences (including our observations)” (Rigakos and Frauley, 2011, p. 247).

(For more on the limitations of empiricism, see Layder (1993, p. 55), Mills (1959, pp. 61–68), and Sayer (2000, p. 11)).

In order to move beyond the empirical, we need a more nuanced conceptualization of ontology; and this is where the model of a ‘depth’ or ‘stratified’ ontology comes into play.

A stratified ontology

is the ontology of critical realist science (i.e., a ‘third way’ of doing science that distinguishes itself from positivism and interpretivism)
suggests reality exists across three analytically distinct ‘levels’
argues what exists goes well beyond what is observable or measurable.

The critical realist depth ontology comprises 1) the empirical 2) the actual and 3) the real.

The empirical is the realm of experience wherein we interact with our physical surroundings and other living beings.

The actual is the domain of realized potentiality, i.e., the world in which all events occur, regardless of whether people witness such events.

The real is the realm of potentiality; it’s where ‘mechanisms’ — the ‘powers’ that cause objects to act or to be acted upon — exist (Collier, 1994, p. 43; Rigakos and Frauley, 2011, p. 249).

The three domains are to be thought of as only analytically (not actually) separate and disconnected; they should be grasped as “overlapping and interacting layers that together make up (social) reality” (Rigakos and Frauley, 2011, p. 249).

As a practical example, in a hockey game,

the empirical is where the referee observes Player A slash Player B, and the referee issues Player A a penalty — the slash occurred and the referee experienced it
the actual is where Player A slashes Player B, but the referee doesn’t see the slash and thus doesn’t call a penalty — the slash occurred, but the referee didn’t experience it
the real is where Player A possesses the capacity (the power) to slash Player B, but this capacity hasn’t been actualized yet — a slash could occur but it hasn’t yet, and, obviously, the referee hasn’t experienced it.

On this model, science must concern itself not only with what can be experienced and measured (the empirical) but also with ‘hidden’ mechanisms and causal structures whose existence must be inferred based on their apparent effects.

(Reread the Rigakos and Frauley quote above regarding power relations if you’re still a bit unclear on the idea that a depth ontology makes it possible for scientists to study real but immaterial things like social structures.)

Another example of something that 1) exists (i.e., is real) and exerts influence on the world but 2) doesn’t possess any material ‘thingness’ and 3) can’t be observed or measured is a social relation.

The concept of social relations recognizes that people inhabit positions or locations that exist and persist independently of the people who (temporarily) fill them (Craib, 1984, p. 22; Hunt, 1993, p. 252).

Examples include the parent-child relation, the landlord-tenant relation, and the teacher-student relation.

Being a tenant with a landlord is an objective position you can occupy; it’s objective in the sense that the relation between a tenant and a landlord doesn’t depend on you in particular for its continued existence. As long as capitalism and private ownership maintain, the landlord-tenant relation will persist.

Social relations are also subjective, though, insofar as they’re a component of everyday lived experience:

“In an important and complex sense social relations are objective. My life and your life are located within sets of relations (of work, politics, domestic, and so on) that are gendered, hierarchic, subordinating, and empowering in ways that (although we may affect their impact) are set up or predate our individual involvement; thus in some important sense they are external or objective.
But social relations are not objective in any sense that implies that they are not a lived part of life within a culture, constructed and lived in and through language and forms of consciousness, as well as sets of objective conditions and sets of social practices.” (Hunt, 1993, p. 252)

It’s crucial for scientists to theorize and investigate social relations, despite the fact that these relations can’t be ‘operationalized’ without destroying what makes them what they are, because these relations both enable and constrain what individual people can do.

Social relations function as a kind of immaterial structure that shapes behavioural possibilities, and immaterial structures are not the kinds of entities empiricist positivism and its commitment to a flat ontology can effectively examine.

Think about it like this:

Kelly, a manager, has the power to terminate employees. From where does her power come? The source of her power is her occupation of the manager position. She couldn’t fire employees if she were no longer a manager. Being in that ‘slot’ is what gives her such power. This is an example of how social relations enable (facilitate) behaviour.
Simultaneously, occupying the role of manager prevents Kelly from taking certain actions. For instance, she might want to provide a recalcitrant employee with one final opportunity to improve his behaviour before she fires him; however, if the company’s termination policy demands that managers fire employees who commit certain actions, and if the recalcitrant employee has in fact committed a prohibited action, Kelly has no choice but to fire him. This is an example of how social relations constrain (limit) behaviour.

These various examples and brief excursions into positivism versus critical realism are meant to demonstrate that core scientific questions — what can we study? how should we study it? what can we explain? — are intimately connected to metascientific explorations into ontology and epistemology.

If we conceptualize the world as perfectly amenable to scientific observation and measurement — as “present[ing] itself to science on a plate, ready cooked and sliced into ‘facts’ [for consumption]” (Collier, 1994, p. 104) — then the use of operational definitions and the commitment to the belief that science deals only with what can be measured becomes justifiable and, indeed, legitimate.

If, on the other hand, we envision reality as complex, stratified across various planes, consisting of immaterial yet real structures and processes, and home to causal powers and generative mechanisms, then a different kind of science is called for.

This is a science that refuses to identify what exists with what can be experienced or measured and that demands more of a relational and structural approach to investigation and analysis.

All of a sudden, then, asking whether a given statement ‘corresponds’ to the world as it really exists, or whether objective social relations can or should be studied independently of subjective intentions and meaning, becomes immeasurably more complex to solve and interesting to think about.

Greater complexity, deeper analysis, and more nuanced ways of conceptualizing the world and scientific study — these dynamics are what the philosophy of science makes possible.

A Few Words About Methodology and the Myth of the Scientific Method

Sixty years ago, C. Wright Mills, now recognized as one of the most influential sociologists of the 20th century, attacked mainstream scientists for their dogmatic dedication to, and use of, what he called the ‘Statistical Ritual’ and the ‘Scientific Method’:

“What [many scientists] have done…is to embrace one [particular] philosophy of science[,] which they now suppose to be The Scientific Method. This model of research is largely an epistemological construction; within the social sciences, its most decisive result has been a sort of methodological inhibition. By this I mean that the kinds of problems that will be taken up and the way in which they are formulated are quite severely limited by The Scientific Method. …
The Scientific Method that is projected [today] did not grow out of…what are generally and correctly taken to be the classic lines of social science work. … Those in the grip of the methodological inhibition often refuse to say anything about modern society unless it has been [filtered] through the fine little mill of The Statistical Ritual. … Much of [science]…has become the mere following of a ritual…rather than…a ‘commitment to the hard demands of science.’” (Mills, 1959, pp. 57, 71–72)

When Mills claims thatthe so-called Scientific Method is merely one kind of philosophy of science, he means to say mainstream scientists have taken one methodology (or metatheory) and treated it as if it represents the only way to do ‘real science’.

This is still the dominant view today.

The story goes like this:

There’s only one true method in science: the Scientific Method.
If you want to carry out genuine scientific inquiry, and if you want to design, execute, and publish legitimate scientific work, you have no choice but to use the Scientific Method.
The Scientific Method is what differentiates science from non-science.
If you don’t use the Scientific Method — if you don’t commit yourself to tests of ‘falsification’, ‘replicability’, and ‘threats to internal and external validity’ — your work will be pseudoscience at best and rubbish at worst.

Was a double-blind, placebo-controlled, experimental study design used, and were all potentially confounding variables and researcher biases neutralized, and were the results replicated by an impartial study team?

No? Sorry! Your results aren’t valid.

This is the attitude of most scientists and laypeople at present.

In fact, what we call the Scientific Method is nothing more than a crystallization of one methodological approach to scientific inquiry — i.e., positivism.

There are, however, multiple methodologies scientists can use to frame their research problems, investigations, and analyses.

We need a bit of terminology here to make things clearer:

Methods are technical instruments and processes scientists use to design and practically carry out their studies; they include procedures such as participant observation, interviewing, and content analysis as well as tools like surveys and software used in the coding of qualitative data or the statistical analysis of quantitative data. Methods are one component of a methodology.
Methodologies, otherwise known as metatheories and research strategies, are concerned with the “logic of implementing scientific methods in the study of reality” (Mouton, 1988, p. 15). Each methodology distinguishes itself on the basis of ontological assumptions, epistemological principles, modes of inference, use of concepts and theories, and status of the outcomes of research.

Methods are about how we conduct research in practical ways; methodologies “shape how we think about our objects of analysis, the research process itself, and the place and role of theorizing” (Rigakos and Frauley, 2011, p. 246).

A growing minority of scientists are rejecting the dominance of empiricist positivism in favour of other methodologies.

The two most popular anti-positivist methodologies are interpretivism (which, by this point, is well known and commonly practiced) and critical realism (which is much newer, less well established, and less popular than either positivism or interpretivism).

(For an overview of positivism, see Delanty (2005, pp. 10–40); for interpretivism, see Delanty (2005, pp. 41–62); for critical realism, see Collier (1994).)

Positivism, interpretivism, and critical realism each advocate and make possible different ways of doing science; this is because they each offer different theories on what science is and should be.

Thus, in this context, methodology, metatheory, metascience, and philosophy of science all generally refer to the same thing: philosophies on the objectives and practices of science; on the nature of reality; and on the possibilities of, and ways for generating, knowledge.

The Scientific Method is nothing more than one philosophy on how scientists should conduct their work. It’s not an objective standard against which all forms of inquiry must be measured. It’s not what differentiates ‘real science’ from ‘pseudoscience’. And it’s not a guarantee that research conducted in accordance with its principles and rules will be accurate, credible, or robust.

The Scientific Method is merely one system for investigating the natural and social worlds, and interpretivists and critical realists would both argue it’s often highly problematic and inadequate.

‘Confounding variables’, ‘falsification’, ‘internal and external validity’, ‘predictability’, ‘reliability’, ‘replicability’ — these and all other elements of the Scientific Method are nothing but theories on how science should be carried out and how research results should be assessed.

Don’t make the mistake of believing they’re objective standards and tests that definitively settle debates in science.

They’re merely one set of concepts from one methodology, i.e., positivism.

Other methodologies equip scientists with other sets of concepts intended to distinguish between legitimate and illegitimate study.

The point is not that positivism is necessarily wrong but that the Scientific Method is by no means the be-all and end-all of science, and it shouldn’t be treated as such.

Keep this in mind whenever somebody insists, for example, that we can’t know whether a set of results are true until they have been replicated.

Summary

In this essay, I’ve argued familiarity with the foundational issues and debates within the philosophy of science (metascience) is necessary to appreciate the nuances of the claims that many philosophers make in their writing.

Many discussions about philosophical and scientific topics — bias, knowledge, objectivity, relativism, truth, etc. — are built on concepts, ideas, and theories that are often taken for granted, i.e., not explicitly acknowledged or explored.

Because these concepts, ideas, and theories are what ultimately make discussions about philosophy and science possible and understandable, I’ve take some time to outline what I consider to be some of the most defining ideas in metascience—ideas you should study if you want to better understand specialized writings on philosophy and science.

Metascience is concerned with the very definition of science. with what science is, what scientists should try to accomplish, and how science operates both in theory and in practice.

The two foundations of metascience are ontology and epistemology.

Ontology is the study of what exists; it tries to offer ways of analyzing the nature of (different forms of) reality and the kinds of relations that persist between entities.

Epistemology is the study of knowledge production; it tries to present ways of thinking about and assessing what can be known, how human beings can generate knowledge, which methods and tools can be used to create knowledge, and what the nature and status of scientific results are.

Whether scientists recognize it or not, ontological and epistemological ideas rest at the heart of all scientific endeavours.

One of the most interesting debates in the philosophy of science is the debate over the nature of reality.

Typically, scientists embrace either a flat ontology or a stratified ontology.

A flat ontology is the defining ontology of mainstream science; it suggests all reality exists on one main ‘level’, and it treats what can be measured as equivalent to both what exists and what is scientific.

A flat ontology leads to empiricism, i.e., the theory that, ultimately, science should be concerned only with the realm of sense perception.

A stratified ontology is the defining ontology of critical realist science; it suggests reality exists across three analytically distinct ‘levels’ (the empirical, the actual, and the real), and it insists what exists exceeds what is observable or measurable.

Amongst other things, a stratified ontology makes possible the recognition and study of things that are real yet immaterial, including social structures and social relations, which exist independently of specific individuals.

Finally, rather than being the objective standard against which all forms of inquiry must be measured, the Scientific Method is nothing more than one kind of methodology (metatheory).

The Scientific Method doesn’t impartially distinguish between ‘science’ and ‘non-science’, nor does it guarantee the validity of research results.

Some scientists make use of methodologies beyond positivism, including interpretivism and critical realism.

There is no one correct way to do science; what science is and how science should be conducted are topics philosophers of science continue to debate.