Philosophical assumptions underlying science


I am a medical student and have been interested lately in the foundations of the scientific research method I have been taught.
I've read that there is in fact no such thing as a unique scientific method. I've even read that science makes no basic philosophical assumptions (or maybe some weak assumptions like "there exists some patterns in the reality"). However, I notice that scientists often speak of "scientific knowledge" in a way that suggests that we indeed possess some scientific knowledge. I find it disturbing because asserting that possessing knowledge (in particular scientific knowledge) is possible is in itself a philosophical assumption (radical skepticism is namely not compatible with this view).
This type of considerations makes me think that there must indeed be some basic, strong assumptions underlying scientific practices.
Can someone recommend me some introductory book dealing with this subject? Many thanks in advance.

EDIT: I have been asked to provide specific examples where scientists speak about scientific knowledge.
Such examples can actually be found very easily. In the field of medicine, for instance, you just have to go to pubmed, open any article featured in the "trending" page and search within the article for the word "know" or "knowledge". Here is what I came with using this method: Here is a quote from this article: "Also, the elevated levels of APRIL may be interesting since APRIL is known to support long-lived plasma cells in its niches". The use of the word "known" is unambiguous here.


Posted 2020-07-30T11:19:52.340

Reputation: 123

why not post this on the physics stack exchange, to see what the scientific community thinks about it? – niels nielsen – 2020-07-30T16:41:22.600

Ideally, science makes only weak philosophical assumptions, but not about "some patterns". They are those required for minimal interpretations of models that connect theoretical and observational languages to allow testing. But that does not mean that science can not make much stronger philosophical claims as a result. Those can be treated like testable hypotheses, albeit in a looser sense. That we possess scientific knowledge is one of them, and is confirmed by the success of scientifically based methods and decisions in practice. That is not an assumption, it is a conclusion, valid or not. – Conifold – 2020-07-30T20:30:09.757


Does this answer your question? What are the assumptions adopted by the scientific community?

– Conifold – 2020-07-30T20:35:36.053

@nielsnielsen I do not see how it would help me to post on the physics stackexchange because my question is not about what the methods used by scientists are but rather why these particular methods are used and what this presupposes in terms of philosophical assumptions. This seems to me to be a typical philosophical question. – user47679 – 2020-07-31T06:14:59.993

@Conifold Thank you very much for your answer. You say in particular that the fact that we possess scientific knowledge is basically confirmed by the success of science. To reach this conclusion it seems to me that we need to rely on a sort of induction principle. Wouldn't it be, then, one additional assumption made by science ? – user47679 – 2020-07-31T06:19:41.613

Not by science. Practical success is largely judged by non-scientists who make the funding decisions. To science this comes as an external test, and so the "induction principle", to the extent that it holds (and it certainly does not universally), is itself confirmed by its success. This is generally true of scientific methodology. On the time scale of individual research methodological principles function as assumptions, but on the larger historical time scale of science they are subject to testing and revision just like specific research claims. – Conifold – 2020-07-31T07:24:38.973

@Conifold Thank you for your answer. Can you recommend me some references dealing with the subject ? I couldn't find any in the thread you mentioned. – user47679 – 2020-07-31T10:55:37.263

I suggest you post it there because it would put you into direct communication with the people who actually use the method, and not just with people who talk about it. As practitioners, they might be able to furnish useful insights. – niels nielsen – 2020-07-31T16:15:36.913

SEP, Scientific Method. – Conifold – 2020-08-01T04:49:07.957

Hi. An outstanding book on scientific method, exemplified especially in the field of medical research, is Ludwik Fleck's Genesis and Development of a Scientific Fact

– Ram Tobolski – 2020-08-01T09:38:10.527

@Conifold I had already read the SEP article, but I do not see where it deals with the assumptions in question. – user47679 – 2020-08-07T06:15:15.833

@Conifold Regarding your comment of 07/31, I am still not sure how to understand it. Do you mean that the underlying principle of each scientific activity would be something like "if the consequences of a proposition (be it a hypothesis in a given science or a general statement about how science should be conducted) have been used with success, then it is worth believing it (compared with a proposition with no practical success)" ? – user47679 – 2020-08-07T06:15:37.387

The point is that there is really no underlying principle, other than that all principles are open to revision, no matter how vague or high level. The only difference is in the time scale on which the revision happens. Even the vague formulation with practical success is, in the end, only provisional. And nothing beyond provisional formulations can be given, or asked for. Their specificity increases with the scope getting limited, in time and subject matter. – Conifold – 2020-08-07T07:01:47.550

@Conifold but if there is no other underlying principles, then how is science different from every other human activity ? How is it even useful ? – user47679 – 2020-08-07T09:28:52.580

There are no timeless universal principles, but those are overrated. To make the question substantive it has to be asked specifically, what distinguishes physics and biology or psychology from other activities today. And then it can be substantively answered. Answers to nebulous contextless questions can only be of a kind, or wrong. You can review history of the demarcation problem for science for various attempts.

– Conifold – 2020-08-07T10:28:04.313

@user47679 The only candidate for "philosophical assumption" in science that you mention is some unspecified attribution of "scientific knowledge". It may help if you add (preferably in the quesiton itself) some specific examples where scientists speaks about "scientific knowledge". It could, after all, be just a manner of speaking. – Ram Tobolski – 2020-08-08T10:09:56.393

@RamTobolski thank you for you comment. I added such an example to my question. – user47679 – 2020-08-08T13:16:53.993

@user47679 Nice example. Yet, how does it differ from ordinary talk? Everyone "knows" e.g. that cars are faster than bicycles. Does this involve substantial philosophical assumptions? In particular, does this challenge the philosophical skeptic? Not typically, I think. The philosophical skeptic does not typically have problems with ordinary language. The issue turns on the meaning of the word "know". (continued) – Ram Tobolski – 2020-08-08T19:02:55.547

The skeptic will typically say, and many will agree, that the word "know" in ordinary language does not carry special philosophical commitments. And the same goes for scientists talk, as long as it is about science, and not about philosophy of science. – Ram Tobolski – 2020-08-08T19:03:27.850

@RamTobolski I am not sure how to understand your answer. I would indeed say that the use of the word "know" in ordinary talk involves philosophical assumptions (although these assumptions are rarely made explicit and many people never really try to make them explicit). In my opinion, thus, the skeptic (in the strict meaning of the term) would be challenged by such use. By the way, this seems to be the whole point of Moore's famous "Here's one hand" argument. – user47679 – 2020-08-08T20:49:03.400

@RamTobolski I think it would be very helpful if you could develop your position (including relevent references if possible), maybe in the form of an answer to the original question I posted. – user47679 – 2020-08-08T20:49:15.870

@user47679 Maybe :) I wanted to ask you for another clarification. "This type of considerations makes me think that there must indeed be some basic, strong assumptions underlying scientific practices" - What would be examples of "basic, strong assumptions"? What would be the difference between a "weak" and a "strong" assumption? – Ram Tobolski – 2020-08-09T12:09:51.630

@RamTobolski I would say that a proposition p can function as a weak assumption if it can be interpreted in such a broad range of ways that it is difficult to find reasons to believe not-p. An example of such proposition could be: "whatever reality is, if it exists, then there are at least some patterns in it". In contrast, a strong assumption takes a greater "epistemic risk" because it is easier to find reasons not to believe it, for example: "reality exists (independently of any subject experiencing it) and is deterministic". – user47679 – 2020-08-09T12:53:24.777



Mario Bunge (R.I.P. 2020) proposed strong philosophical assumptions underlying the process of making science. I would recommend La ciencia, su método y su filosofía, 1960, Mario Bunge (don't know of english translations). Probably there's a lot of more from him in english, he was quite prolific, a great philosopher of science and an amazing epistemologist, and I'm far from reading 2% of his work. Following, what I remember from reading him, which could answer your question.

This is, in simple words, what the scientific method is about (at least, an interpretation of Bunge's ideas, I tried to be as precise as possible).

First, knowledge is more or less a model of reality that is built in order to increase the probabilities of survival. Second, knowledge is subjective. Third, scientific knowledge is just some type of knowledge, which is build using the scientific method.

"Using the scientific method" could be subject to multiple debatable interpretations. One of my best teachers (using the aforementioned book) stated that the scientific method just reduces to reach the optimal level of objectivity (which has the goal to allow the communication of knowledge, in order to be useful for others, in order to survive). What does it mean?

Considering the previous three statements, the fact that knowledge is subjective means this: you know the sky, I know the sky, but the concept sky is not objective: it is the result of multiple experiences of each one of us; then, it is subjective for each one. But take into account that such experiences are necessarily different (otherwise, it would imply that you and me are the same individual). Therefore, at least, our knowledge of the sky is a shared subjectivity. If we talk about it, we can assume that we agree on multiple facts about the sky, but by doing that, we will not reach objectivity. That's just sharing inter-subjective knowledge (my phrasing, not Bunge's). Religious knowledge is usually inter-subjective.

When would we state that some knowledge is objective? When can we trust that communicating knowledge is safe? The answer is this: the moment you and me apply a method that both agree on, and that make knowledge as objective as possible, so to be communicated, and used for survival.

That's precisely the scientific method. A convention, an agreement on how to reach an optimal level of objectivity. An excess of objectivity would be undesirable (we don't need to describe in excessive objective detail where are the atomic limits of solids in order to agree that adding 1+1 solids results in 2 solids). A lack of objectivity is undesirable as well (I could affirm that God exists because I had a dream, and you should agree). The scientific method produces a category of knowledge that we agree on by convention, which can safely be communicated to others that need it. As you see, there's a lot of issues to agree on. That's the philosophy of science, which could be a huge amount of propositions and rules.

The philosophy of science must not be misunderstood with the knowledge that is product of science. The philosophy of science is our agreement on how to develop knowledge that aims towards our survival. The scientific knowledge is such knowledge in itself.

See for example the historic development of the thermodynamic laws. The first three laws were created and nobody noticed that there was a huge subjectivity embedded on them: the concept of temperature. Temperature is a feeling, not a physical concept. But that was quickly solved, just by adding a previous law (the zeroth-law), which defines in objective terms what is temperature.

That is the process of the scientific method. Not necessarily what common literature describes (observations, hypothesis, testing... etc.). So, it is clear that the scientific method does not reduce to a short and rigid set of rules, but rules are useful in multiple contexts.

Bunge makes a deep analysis not only of the complexities of such process but also of the problems of communication, the epistemic issues (how to deal with multiple types of knowledge?), the pragmatic consequences (a discipline would be characterized from three dimensions: science, technique, art), etc. Highly recommended.


Posted 2020-07-30T11:19:52.340

Reputation: 2 572

The idea knowing 'reality' directly correlates with survival doesn't hold water - see Donald Hoffman

– CriglCragl – 2020-08-19T23:19:19.923


@CriglCragl See this counterargument to Hoffman: Manolo Martínez (2019), "Usefulness drives representations to truth: a family of counterexamples to Hoffman's interface theory of perception", Grazer Philosophische Studien, 96(3), 319–341.

– Big Mac – 2021-01-26T03:40:19.873

1@BigMac: Hoffman is not an anti-realist, he simply says we must apply reason to perception, rather than assume reason & evolution (perception or sensory experience) must converge. For instance, repeat observations, or seek consilience from multiple senses. This amounts to cautioning against cognitive bias. – CriglCragl – 2021-01-26T18:49:54.463

@CriglCragl Thanks. That's true enough as far as it goes. Much more to say about this, of course: e.g., the whole field of evolutionary epistemology. A big problem in the definition of science in the answer above is that much of science has no apparent survival value. – Big Mac – 2021-01-26T22:27:31.340

@BigMac: Indeed. But the interests of our minds (eg global workspace with a strange-loop generated by self-reference) can diverge from the 'interests' of our genes, the unit of selection. Especially, where an additional layer of selection occurs, which by multi-level-selection principles, must coincide with individual/gene benefits, until it begins to provide additional, greater benefits - slime moulds, social insects, and the human memesphere, or language, able to confer benefits even from non-breeders innovations/behaviours, onto the wider population (eg kin of non-breeders). – CriglCragl – 2021-01-26T22:39:55.000


Talk of scientific knowledge is indeed presumptuous (though by no means unjustifiable), but in a scientific context I'd suggest to dismiss it as rhetorics mostly. After all, while scientists go for theories that are (somehow) best supported by the evidence, they lack any justification for ever claiming to have, in fact, discovered the truth. (At any rate, it is not their business to do so.) For a recent introductory book you might want to read Lee McIntyre's "The Scientific Attitude" (2019).


Posted 2020-07-30T11:19:52.340

Reputation: 170

How is reproducible empirical evidence "lacking any justification for discovering the truth"? – Cell – 2020-07-30T12:42:24.763

Well, I was not saying that scientific claims lack justification if that is what you mean; clearly that is not the case. And by no means did I want to suggest that scientists could not ever be right about the evidence per se (though that is something to be discussed). Rather, I meant to say that there are no scientific resources for evaluating a theory beyond the empirical evidence it purports to explain; and such an evaluation is obviously bound to be defeasible. – Turtur – 2020-07-30T13:07:06.370


A solid history chronicling how traditional philosophy of science and scientific positivism evolved during the 20th C into post positivism and anti-realism, bringing about the (to some extent) displacement of the philosophy of science/knowledge (by blurring boundary between epistemology and ontology), and its replacement with the sociology of science/ knowledge, and the extent to which the project was and was not warranted by some of the factors mentioned in your post) is John Zammito’s A Nice Derangement of Epistemes: Post-Positivism in the Study of Science from Quine to Latour. (

Zammito contends that while some skepticism about empiricism, positivism, and the claim that scientific theory choice is premised always and exclusively upon rational bases, is warranted, he laments that three "hyperbolic dogmas" of anti-empiricist theory are primarily responsible for the unfortunately and unwarrantedly radical epistemological "shift" described above, and which likely informs your question about the "assumptions" which underlie scientific practice: the theory ladeness of perceptions/observations; the underdetermination of of scientific theories by evidence, and the incommensurability between competing theories [the claim that disparate incommensurable theories equally well describe/explain phenomena, or "reality"]. The bookss footnotes constitute an impressivly extensive bibliography of the primary texts in the area.

Addendum: It occurred to me that before reading Zammito (whose goal is to moderate and contextualize the trending "radical skepticism"), you may wish to first actually sample first hand the skepticism about traditional scientific empirical rationality upon which the more radical skepticism is based, by reading an introductory text or two. For instance, the anthologies Challenges to Empiricism (Edited by Harold Morrick in 1972), or Post-Analytic Philosophy (Edited by John Rajhman and Cornel West in 1985) are quite good. Or Simon Blackburn's 2005 On Truth: A Guide, or anything by the philosopher of Science Hillary Putnam, or the much more radical Richard Rorty.

Moreover, as suggested in another answer, auditing a "science studies" class in the humanities or social sciences might also be a good idea. These classes are generally taught by professors that were raised on and are steeped not in science, the philosophy of science, but rhetoric. Maybe a class in the "Sociology of Science" or "Science and Technology Studies" (see here:

Aside: Finally, given the use of the term "radical skepticism" in your post, you may also find this post of interest: How do philosophers respond to global skepticism?.


Posted 2020-07-30T11:19:52.340

Reputation: 1 666


If you're attached to a university, I'd suggest you audit a class on the Philosophy of Science over in the philosophy or humanities department. You'll get a quicker and more solid overview of the material than you will through self-study. Don't worry: the professors over there are generally pretty genial, and their classes are never over-crowded, so likely they'll welcome you easily.

I often find it helpful to point out to people that the 'Scientific Method' (to the extent that we can talk about such meaningfully) isn't a method of action or practice, though it is often portrayed that way. It is a method of reasoning, a particular way of looking at and thinking about the world that can be (and is) applied in a multitude of different ways in different fields and contexts. The philosophical assumptions are the following:

  • That the world (universe, material reality) unfolds in a systematic (law-like) manner
  • That is is possible to model or describe this systematic unfolding using abstract theoretical structures (often mathematical, though not as a matter of necessity)
  • That these models/descriptions are never perfect, and can be developed and improved by carefully comparing and contrasting them to what we perceive in the world around us

This process of reasoning is confounded somewhat by the philosophical recognition that we cannot always trust the prima facie evidence of our senses — e.g., on face value it appears as though the sun rises, though we have learned better — so there is sometimes an unpleasant tension between our ingrained predispositions and the systematic models we develop. But that is the process of scientific reasoning: to rationalize the one against the other until they are mutually coherent and consistent.

Ted Wrigley

Posted 2020-07-30T11:19:52.340

Reputation: 9 139

"That the world (universe, material reality) unfolds in a systematic (law-like) manner." This is a purely speculative metaphysical assumption which cannot be demonstrated in any meaningful way and which contradicts the accepted 'scientific' dogma that the human perception of the sensible world cannot be trusted to deliver any truth value. It represents a superstitious belief in a complete unknown entity, laws of nature, that delivers certainty through a transcendental process. It is no different than belief in an anthropomorphic god/entity. Your second assumption is untenable. – None – 2020-08-20T14:37:29.943

@CharlesMSaunders: Whether or not it's correct (and whether or not you like it), those are the philosophical assumptions of science. And if the first principle is speculative, it's a speculation borne out by an imponderable mass of empirical data. I'm not really certain what point you'r trying to make, aside from generally rejecting the self-evident. Could you clarify? – Ted Wrigley – 2020-08-20T15:02:20.200


Science shifts the ground beneath our intuitions about ideas like truth, and knowledge, and causality. Truth and knowledge become tentative, contextual, subject to change. I would point out that has been true even geometry, or logic, where non-Euclidean mathematics & Godel's theorems have fundamentally shifted what we thought were fixed results. But like we didn't bin Newton's theory of gravity, we never binned the previous maths & logic, just set them in a bigger context. Even seemingly fundamental truths & knowledge have become subject, always, to that, not being dismissed but included in a larger whole. Given Hume's problem of induction, causality becomes about a narrative grouping, rather than (in general anyway) a provable necessity.

I would defend the framing "science is what scientists do". This can seem terrifyingly arbitrary, or subject to unlimited amendment. But I would say rather it is a language and a culture, which has grown into it's current sophistication like a language does, or a species.

Rather than being founded on assumptions, I would look to Durkheim's idea of sacred values binding together moral communities, to understand how this community is what it does. To challenge core values of a culture, like say habeus corpus in Britain, or the right to free speech in the USA, would be to challenge the cohesion of these cultures, and the sense of shared identity renewed in the practice and stories about themselves by which they propagate. Publishing theorems results and techniques in reputable and accessible form which will be stored, is a key value - the dispute, feud even, between Newton and Liebniz, and between their followers, over the creation of calculus, is as much responsible for this as any core assumption. Internationalism, the idea there is one scientific community, and each discipline expects to be able to reconcile differences - Soviet agriculture, and Nazi physics, were examples of attempting local cultures of scientific epistemology. The international condemnation and risk of scientific sanctions towards Korea and China over reckless human germline experiments are another example, as are limits on weapons development, it's not just about finding shared epistemology, but also safety and morality, to remain part of the global scientific community. A new 'branch' of the scientific community could start, it could cause a schism or fragmentation, or become the new mainstream. This has happened, often, in nearly every discipline, and is a big part of how the scientific method develops.

Skepticism and settling arguments by evidence rather than authority are key values from the earliest days of science, but I suggest even they are witheld sometimes, with heuristics like 'naturalness' in physics which are unprovable guides, or avoiding engaging with unserious or malign critics of climate science, respectively for instance.

I recommend How The Laws Of Physics Lie by Nancy Cartwright, even though it's focused on physics. I'd describe all other sciences as largely similar, with more heuristics. Her perspective is we try to make valid or valid enough abstractions, to make the scientific work tractable, and correspond to the system with the greatest simplicity/efficiency we can. These systems of abstractions are always limited by how true the assumptions behind them are. And the world always has the last word.

The really fascinating thing about how science progress is not I think evidence gathering to decide between models, but hypothesis generation. Popper made a strong case that these cannot be generated 'mechanically' from the data. It requires creativity, insight, intuition, innovation, things which are arguably beyond science. We should celebrate and communicate this to non-scientists, it is where the thrill and revelations of scientific discovery occur.

I would situate the intelligence of science, in our collective 'eusocial' or hive-like intelligence. Which is not to say it suppresses individual insights or creativity, but like the sophisticated cooling of termite mounds or bee hives, generates something beyond any individuals powers through a set of ways of interacting that foster emergent behaviour that benefits the whole community.


Posted 2020-07-30T11:19:52.340

Reputation: 5 272

@CriglCragl- The practice of science being often benign or beneficial is not the problem. It arises from the dangerous downsides which arise and continue to. People's lives in major European cities have been recognized to have been shortened by up to one and a half years. Only the scientific community in the person of the philosopher of science can attempt to correct this egregious misapplication of combustion. Your dreamy and idyllic view of the scientific process marks you like the other respondents to this question as 'sleepwalkers'.. Science will always remain fine, but not infallible. – None – 2020-08-18T12:31:55.527

@CharlesMSaunders: I really don't get your point, or it's relevance to the original post. – CriglCragl – 2020-08-19T23:16:01.483

@CharlesMSaunders: Ok, I get your point, now I read it again. It's like blaming printing presses for the witch-panics. Pamphleteering was definitely part of it. Do we then have a 'dreamy idyllic sleepwalker' view of being literate? Printing presses demanded more critical thinking, and universal education. The advances of science created unseen problems, and the answers to them invariably are in doing more science, better - eg codes of conduct about human germline experiments in order to fully access the community (eg conferences, standards-setting bodies) as I put in my answer. – CriglCragl – 2021-01-26T22:53:20.393


The best book I've read to get a deep grasp from an introductory starter level is Lewis Wolpert's "The Unnatural Nature of Science." Buy used--it is going at a real premium new the last time I looked. He just passed away January 28 of this year at age 93.


Posted 2020-07-30T11:19:52.340

Reputation: 1