I think a combination of science and philosophy alone leads to closer to truth.

Why is it that it is wrong to think that science alone leads to truth? Isn't empirical truth the only one we can be sure about? If there is any absolute truth at all, isn't it to be uncovered through the scientific method?

  • 7
    There is also a priori knowledge which is what makes empirical knowledge possible, as Kant held: "There is no danger of [the possible discovery that there is no a priori cognition at all]. It would be tantamount to someone's wishing to prove by reason that there is no reason. For we say that we cognize something by reason only when we are conscious that we could have known it even if we had not encountered it thus in experience; hence reason's cognition and a priori cognition are one and the same."
    – user3017
    Commented Feb 26, 2018 at 15:01
  • 4
    You said that science leads to empirical knowledge: this is true. We have also mathematics: it seems not empirical, but for sure it is a huge field of "knowledge". Commented Feb 26, 2018 at 15:16
  • 4
    Not clear... I read "absolute truth" as certainty, indubitable knowledge. If so, no absolute truth at all. Commented Feb 26, 2018 at 15:40
  • 2
    See also the fact that the principle of scientific induction is exactly as self-justifying as its negation ("if it's happened this way in the past, it won't happen this way in the future"). lesswrong.com/lw/s0/where_recursive_justification_hits_bottom Commented Feb 26, 2018 at 20:06
  • 8
    What is truth? I'm not kidding. What is truth? Can you define that concept using science alone? If you're interested in this line of thinking, I can turn it into an answer, but it would probably be a better answer if I could get your opinion on the answers to those questions and tailor the answer to fit them.
    – Cort Ammon
    Commented Feb 26, 2018 at 20:26

10 Answers 10


The problem with scientism is that it's generally philosophically incoherent. Examine your own statements of scientistic dogma:

Isn't empirical truth the only one we can be sure about? If there is any absolute truth at all, isn't it to be uncovered through the scientific method?

Laying aside the thin veil of casting these as rhetorical questions, neither statement is an empirical truth, nor has been uncovered through the scientific method (if you disagree, describe to me the experiment that confirms them). They are therefore self-undermining statements of belief. You could have instead said "The only truths I accept are the ones that have been empirically confirmed," or "The scientific method has been the most valuable method of intellectual inquiry for the human race," but those statements rescue themselves from self-negation only at the price of making their value judgments more explicit.

There isn't necessarily anything demonstrably wrong with affirming science as your own personal belief system, as long as you understand that is what you're doing. But the belief that science itself confirms scientism, or is even capable, structurally, of confirming scientism is incorrect. That's not the kind of thing it was designed to do --those are exactly the kinds of pronouncements it withholds.

  • 13
    Also, if a scientist discovered a better method of learning than the scientific method, they would start using that. 500 years from now history books would say "For many years, research was predominantly done via the scientific method, an elementary form of research using experiments and drawing conclusions." And if that sounds ridiculous, well, consider all the work that was done without the scientific method, especially a formalized one.
    – corsiKa
    Commented Feb 26, 2018 at 18:30
  • 12
    @CriglCragl No, the scientific method is a standard series of steps used across all science. Define your hypothesis, design an experiment that tests that hypothesis with predicted results, perform the experiment using controls, record observations, and compare observations with predictions. Then the process begins of explaining deviations between predictions and observations. It's not a metaphor, and the fact that it isn't common knowledge what the scientific method is is a failing of our school systems.
    – corsiKa
    Commented Feb 26, 2018 at 20:41
  • 28
    @MartinArgerami I promise you I have a marvelous retort to your baseless insult that this comment is too narrow to contain.
    – corsiKa
    Commented Feb 26, 2018 at 21:56
  • 11
    @CriglCragl While it's certainly true that the implementations of each step are different across fields, that the steps exist across fields is not in question. There are no fields of science that lack hypotheses, none that lack experiments, none that lack observations, and none that lack comparing results to predictions. (There are individual scientist who lack them, and their work is filtered out via peer review, assuming infallible humans doing the checking.) Further, I fail to see how even if it were true that they were that the scientific method is metaphoric in any way.
    – corsiKa
    Commented Feb 26, 2018 at 22:58
  • 10
    @CriglCragl "a lot of cosmology is done with simulations": this is true. But there would be little (scientific) point in those simulations if they were not based on previous observations and designed to make predictions in order to test hypotheses. There are many steps involved in studying something scientifically, and not all of them by themselves involve every component --- but the term "the scientific method" refers to the whole thing, when you put it all together.
    – N. Virgo
    Commented Feb 27, 2018 at 4:50

While the notion of reproducible results under specific conditions does redeem science from having its methods be entirely faith-based, initial presuppositions are always necessary in order to form a coherent paradigm. So, if you start with one (or many) axioms, then you've imported several implications from the axioms into the paradigm itself, which means that it does include some form of untested or "untestable" belief.

You might enjoy (or be driven mad by) this podcast between Sam Harris and Jordan Peterson on epistemology.


Your question is confusing and badly written. Your first sentence:

I think a combination of science and philosophy alone leads to closer to truth.

contradicts your second paragraph:

Why is it that it is wrong to think that science alone leads to truth? Isn't empirical truth the only one we can be sure about? If there is any absolute truth at all, isn't it to be uncovered through the scientific method?

Your position is that there is a process called "science" that leads to something called "empirical truth" by the "scientific method". There are several problems with this argument.

Let's consider this process that you say leads to truth. You have some set of facts and some rule that you apply to them that produces truth somehow. The results of that process are only correct if your facts and rules are correct. But then you have to know that the facts and rules are correct or your conclusions may be wrong. At this point you face a problem. You either (1) say the facts and rules are correct by fiat or (2) say the facts and rules are shown to be correct by some other process. If you choose option (1) then you're a just a guy saying stuff on the internet. If you choose option (2) then you have to show the truth of the facts and rules used in your new process: process 2. You then have the same dilemma with process 2, and with whatever process you use to show process 2 is correct, so you get an infinite regress. There is no process that guarantees truth or correctness, and that includes science.

There is a further problem with your position. To decide what is true according to your position, you have to distinguish between science (true) and non-science (false). But whatever process you use to make that distinction isn't science since you can't tell science from non-science before you have it. So your worldview can't consist solely of science.

Philosophers say a lot of stuff about induction and justification of science, and almost everything they have written doesn't answer the problems I have pointed out above. One philosopher whose work does not suffer from this problem is Karl Popper. Popper proposed that all knowledge is created by spotting a problem, guessing solutions and criticising guesses to eliminate bad ideas. This doesn't require any process of showing ideas to be true. If you have two conflicting ideas, then at least one of them is false. In two books "The Fabric of Reality" and "The Beginning of Infinity" physicist David Deutsch elaborated Popper's ideas pointing out that when you set up an experiment you need an explanation of what is happening in reality in the experiment to interpret its results and without such an explanation your 'results' are worthless. This means that the empirical truths are all guesses too: they are not a secure foundation for certainly true ideas. You can improve your ideas indefinitely, but you can't have proof or certainty. For a guide to Popper's writings, see


  • 1
    Science is the search for facts based on observations with a given degree of uncertainty from which predictions may be made, not truth, and nothing more than that.
    – Mazura
    Commented Feb 27, 2018 at 19:44
  • 3
    @Mazura No. Science attempts to explain how the world works, not just make predictions. For example, current scientific theories make statements about events we can't observe and will never be able to observe, like what's going on in the core of the sun right now. In addition, you can't make predictions without explanations that involve unobserved events. Two samples of a chemical are the same chemical if they contain the same kinds of atoms even though you don't observe the atoms.
    – alanf
    Commented Feb 27, 2018 at 20:07
  • "Science attempts to explain how the world works" - Not really. What you're talking about are scientific models, and while they are a long lived and important tradition in science, it's not at all "what science does." "In addition, you can't make predictions without explanations that involve unobserved events. " In the sense that predictions are by definition about things that have not yet been observed. But like, you can make predictions about cars, which obviously you've seen before. "even though you don't observe the atoms." -> symantics
    – industry7
    Commented Feb 27, 2018 at 20:38
  • @alanf - That's a conflation of what science is and what scientists do. But yeah, we're way into semantics by now ;)
    – Mazura
    Commented Mar 1, 2018 at 23:53
  • @mazura The point of science is to find truth. You have to judge whether a theory or observation is good by some set of standards and one of the standards is whether the theory or observation could correspond to reality, i.e. - whether it could be true.
    – alanf
    Commented Mar 2, 2018 at 13:24

Chris Sunami's anwer is excellent. I would like to add to it by pointing out the fact that "the scientific method" isn't as well defined as you may think, and defining this method is in the realm of philosophy.

While some people argue that you should just "Shut up and calculate", others (notably Nancy Cartwright) would argue that the Scientific Process isn't just the building of a mathematical model, but also the building of a philosophical model and understanding how the two relate (explained quite well in this book).

Technicalities as to how the scientific method works means that you not only have divisions between philosophers arguing on the subject, but even people who say they believe in "scientism" also have contentions on. For example, you have the split between the people who think that all scientists are discovering something about reality and those that disagree that anything they discover is "real", since it is not the final Fundamental Unification theory.

As you study more science, you discover that interpretation of theories (and choice of the theories you deem worthy to interpret) cannot be based on any ground truth implicit to the theory without resorting to some philosophical principles.

  • 2
    Au contraire - if the scientific method only took into account whether something "works", we would still be using epicycles to describe the solar system. When Heliocentrism was adopted by Copernicus, Galileo and their contemporaries, it didn't produce concrete results to the same accuracy as Ptolemaic astronomical models. A philosophical paradigm shift was needed, where it was pointed out that it was far more aesthetic to have simpler circles around the sun. When Kepler's laws allowed a better precision in astronomical predictions, heliocentrism finally caught up. Commented Feb 27, 2018 at 23:38
  • 2
    But a complicated enough system of epicycles could also describe the solar system. Of course, it is obvious that we must reject theories that do NOT describe reality at all. But even that is contended - some proponents of scientism would reject the scientific results obtained from biology and neuroscience, because they haven't been achieved through a study of their parts beginning from physical theories. And how do you pick between two theories that have the same results but different ontological implications? Most scientists resort either to Occam's razor (which is a philosophical concept)... Commented Feb 27, 2018 at 23:42
  • 1
    ...or to the aesthetics of a theory (also a philosophical concept). Science used to be called "Natural Philosophy" and the only reason scientists would distance themselves from philosophy is if they are too scared to defend the philosophical position they hold while they work - as can be seen in Tegmark's ignorant position presented in his "Shut Up and Calculate" ( arxiv.org/abs/0709.4024 ) Commented Feb 27, 2018 at 23:45
  • 1
    @Graham I'm not sure you read what I wrote, you seem to be avoiding all the points I made. You write about Brahe, but Brahe had a Geocentric model of the universe with the planets orbiting the sun, which orbited the earth. This raises another interesting point - the universe can be described from any reference frame, and if it is described from the reference from of a stationary earth, the universe rotates around it, and the solar system moves somewhat similarly to Brahe's model. This can be done with any modern theory too, including General Relativity. The fact we say the Earth orbits the Sun Commented Feb 28, 2018 at 11:39
  • 1
    "to the extent that philosophers argued whether observations of physical objects through a telescope were actually real." This is about the proposed introduction of a thoroughly new tool of which around the time the Galileian/Dutch variety had small field of view, no real image (i.e. you cannot produce the image on a screen) wheres Kepler's variety had larger field of view and real but inverted image. And with the means lens production, they probably all produced blurry and distorted images. Given this, I consider it proper science to discuss reliability and limitations of these tools. Commented Apr 2, 2018 at 16:08

Sorry, I had to make my comment into an answer. In short:

Did you just ask an epistemological question wondering whether we need epistemology? The Vietnam generation had a similar paradox regarding peace and reproduction.

  • 2
    I understand (and like) the sentence that you've written in bold, but I'm a little confused about how the sentence that follows it is linked with it. That it is linked appears to be the case because you write 'similar'. It might be worth explaining it in a little more detail. Commented Feb 27, 2018 at 15:09
  • 1
    @jpmc No, I was saying that fighting for peace is like fucking for virginity (is like using epistemology towards a pure nature-science based thinking). Commented Feb 28, 2018 at 6:09
  • 1
    @PeterA.Schneider Never heard that, and wish I still hadn't. Apparently, a significant portion of your readers don't know it, either.
    – jpmc26
    Commented Feb 28, 2018 at 6:35
  • 2
    @jpmc26 That the main takeaway of that era has apparently fallen prey to the oblivion may explain current U.S. policies. Commented Feb 28, 2018 at 8:31
  • 2
    @PeterA.Schneider I also do not understand what is paradoxical of ending a war via fight. You might be influenced by selecting a war with which you do not agree. Take WWII instead - we did end the WWII by fighting Hitler. We did end the Napoleon wars by fighting Napoleon. There is not anything paradoxical. The statement "fighting for peace is like fucking for virginity" is a pure nonsense, it's only a witticism. Commented Mar 1, 2018 at 17:03

Many excellent answers here, but I just want to address this one question in your post:

Isn't empirical truth the only one we can be sure about?

No. Not even close. Empirical confirmation is a form of inductive reasoning. We run many tests, and if they consistently pass, we accept the hypothesis as true. But inductive reasoning has a fundamental weakness: it is never absolute. It is always, always, always susceptible to new information coming to light.

Incredibly famous examples are the rise of quantum and relativistic physics. After several hundred years of studying Newtonian physics, we started to discover situations it could not explain: the orbit of Mercury and why atoms didn't collapse from constantly radiating electromagnetic waves are good examples. Yet up until this point, the empirical evidence suggested Newtonian physics were fully correct. And even though Newtonian physics isn't really wrong (It's really good at solving problems at the scales humans normally work at.), its incompleteness means that its predictions are not absolutely true. Quantum and relativity are also incomplete: quantum cannot predict cosmological scale events, and relativity cannot predict subatomic events.

In other words, embracing empiricism means you're never completely sure your existing knowledge applies to any new situation you encounter. You never know if your empirically confirmed model is missing some important factor you've never needed to account for before. Empirical evidence is only a useful tool in gradually reducing the flaws in your understanding; it can't eliminate all of them.

  • I would also disagree with the OPs statement - because it is a Straw Man. Scientific experiments do not deliver the truth - whatever that might be; rather, they produce a model that is confirmed for the range of values studied in the experiment. Study the motion of the planets with a telescope and stopwatch and you get Newtonian mechanics. Move to density scales of a Black Hole and you get Relativity. Try again with sub-atomic particles and you need Quantum Gravity. None of them are the Truth. They are just increasingly precise models that depend on the scale of the experiment. Commented Feb 28, 2018 at 11:12
  • @OscarBravo It depends on what you mean by "the truth." Someone who emphasizes empiricism would probably be satisfied with that approximation as "truth," especially since it's often information that can inform decisions about how to achieve our goals. But yes, if by "the truth" you mean something like "every aspect of objective reality" or "all facts that exist," I agree.
    – jpmc26
    Commented Feb 28, 2018 at 11:25
  • You are presupposing that some kind of objective truth actually exists. If we restrict ourselves to the physical universe, it could be there is no truth. There might only be increasingly more refined models, each of which breaks down at some smaller/faster/more energetic scale. This could go on forever... Commented Mar 1, 2018 at 8:35
  • @OscarBravo I do believe there is an objective truth, but my comment itself presupposes nothing. It merely discusses a couple of different viewpoints and their consequences with regard to this topic and notes that your comment didn't wasn't entirely clear what it meant. As for, "this could go on forever," of course. All systems of logic presuppose a set of axioms, which can be challenged endlessly. That's nothing new.
    – jpmc26
    Commented Mar 1, 2018 at 9:06

Why is it that it is wrong to think that science alone leads to truth?

Simple counter example: the problem of induction. We have no reason to believe that inductive reasoning is valid, except for our past observations that inductive reasoning had proven useful to us. Extrapolating these past successes to claim that "induction will continue to be valid", is itself an argument using inductive reasoning, making it a circular argument.

In other words, there's no non-inductive proof that proves that inductive reasoning is valid. Science is all about classification and discovery of the world through observation and inductive reasoning. There's no way the scientific method could be applied to prove that the scientific method is valid.

Of course, these are extremely skeptical view points. You could never live your life under the belief that induction isn't valid, it just isn't possible. The mere existence of human memory, learning, education, etc. is reliant on the idea that the mistakes of the past can be learned from to better inform decisions of the future. Nonetheless, from an abstract/philosophical perspective, the issue remains.


Scientism is also scientifically wrong in so far as it is (at least partially) a scientific question how homo sapiens acquire knowledge. "Through science alone" is not a good scientific hypothesis for how knowledge is obtained.


I am not decided on the issue. What I will do here is attempt to work out a consistent and non-absurd scenario in which moderate empiricism is true and we come to know it. After developing an account, I will see if it holds water. I'll be happy to receive improvement suggestions. Note that I am at best familiar with the discussion.

I. Suppose moderate empiricism is true. Here is a relaxed, impressionistic attempt at formulating it: Our only source of justification for matters of fact (and not of definition) comes from repeated interaction with reality. I suggest that a creature inhabiting this scenario could conceivably learn about its brain, its reasoning capabilities, and features of the world it inhabits in general. Then it could infer moderate empiricism as the best explanation for what it has learned.

Of course, the creature could only do that if it knew (or was epistemically justified in believing) that its explanatory judgments are reliable, i.e., that the theories they judge to be great explanations are very often true. How could this occur?

If there couldn't be any simple set of experiments to confirm them, then the creature could learn of the reliability of its own explanations by observing that they very often lead to predictive theories.

Quine proposed something similar to account for how we are justified in believing certain fundamental statements in mathematics, logic, epistemology, and metaphysics, but not justified a priori. His point is that these statements are central to the way we reason when interacting with reality, and that since we interact successfully with reality, it must be because they reflect something of reality. The argument given for the creature's explanatory practices above follows a similar trail.

Finally, the creature's modes of explanation could conceivably lead it to discover moderate empiricism. It is difficult to work out how, since it would require spelling out what kind of evidence would call for moderate empiricism as the best explanation. However, all I needed to show is that basic tools of reasoning could be justified in a moderate empiricist scenario. This opens the possibility of arguing we do live in this scenario. Have I at least suggested how such showing could be done?

II. The problem with the above account is that it does not seem the creature could get off the ground. The creature in question would need to employ a set T1 of reasoning tools in getting from (a) its immediately-known empirical knowledge that, by using a set of reasoning tools T2, it has come to be a successful predictor and controller of reality, to (b) the conclusion that tools T2 are truth-conducive. Having established 'b', the creature would be many steps closer to being justified in the conclusions it draws out by employing T2.

Yet, a b-like statement is necessary for having epistemic justification in employing any set of reasoning tools. Thus, a b-like statement for T1 would need to be obtained (probably from an a-like statement that mentions it). Yet, that would require a previously-justified set of reasoning tools T0. Since this obviously leads to an absurd regress, it follows that the creature has no justification for any of the tools of inference it employs, and so no justification in drawing out its conclusions!

So it seems it cannot learn about moderate empiricism after all. It cannot learn about anything it does not immediately perceive. Perhaps it can barely do any interesting empirical psychology (which Quine thought was integral to empiricist epistemology), since it cannot immediately perceive many of the entities and mechanisms postulated in good psychological theories.

III. Could the above conclusion be averted for the creature? I can imagine three possibilities.

(1) Perhaps the tools' reliability can be discovered non-inferentially. Maybe just observing them work in varied situations, plus the fact that they do work, is already direct acquaintance with their truth-conduciveness. It needn't be inferred using tools of reasoning, and so regress does not surface.

(2) Perhaps a moderate coherence theory of justification is correct. Given a web of beliefs full of direct observational knowledge of the world, and given that the web is stratified by tools of reasoning, the mere fact that it managed to stay consistent, let alone coherent, is already an immense signal that such tools are correct. Coherence theories have been developed, AFAIK, in which an agent can acquire justification from coherence without being previously justified in believing the coherence theory. (The agent needn't even know that its set of beliefs is coherent or consistent, given objectively correct standards of coherence and consistency.)

(3) Maybe a holistic theory of justification is correct, and a creature's belief-system obtains justification as a whole when the creature employs it in engaging successfully with reality. This way, the reasoning tools the creature uses could acquire derivative epistemic justification, since they are an integral part of that web of beliefs.

The regress of section II might be avoidable given '1', '2', or '3'. What do you folks think?

  • "since we interact successfully with reality, it must be because they [the fundamental statements] reflect something of reality" or, mabe, we call that subset of the world [better term anyone?] reality which is reflected/accessible via this system of fundamental statements. Commented Feb 27, 2018 at 16:00

If there is any absolute truth at all, isn't it to be uncovered through the scientific method?

No. The scientific method focuses on observable "objective" truths and presumes that the set of observable truths and truths yet to be scientifically proven are the entirety of "absolute truths". This results in modes of thinking and mental frameworks that eventually result in deviations from actual underlying truths and can be observed in science's failure to accurately and consistently predict and anticipate observable outcomes in marginally complex systems.

For example, a byproduct of this mode of thought is that characteristics and qualities of a thing are contained within that thing. For example, elements have innate qualities, animals have innate qualities and... people have innate qualities.

But what's the innate quality of an "In-N-Out" burger? Or a Chick-Fil-A sandwich? Is it good b/c of the balance of saturated fats and salt? Or is there some element of quality that exists between the observer and the observed? Two individuals may observe a favorable taste in either. Two hundred thousand may observe the same. But apart from surveys and statistics, science is, at best, a blunt tool in measuring the "burger-ness" of the In-n-out double double or Chick-Fil-A Deluxe Sandwich.

Yes, science has broken down the beef patty to the extent that we can recreate a plant-based version to a surprising level of similitude (Impossible Burger yum!). But what about presentation? What about the employee's level of service? What about the supply chain that delivered that burger? How do those factor into this feeling of quality that some of us feel towards In-N-Out (or pick your favorite restaurant chain)?

This is where science starts to strain itself as it tries to break down each of those component mechanisms and factors and back into some quantifiable sum of the parts. Whereas we, as observers, can simply and factually say whether a particular restaurant is good, ok or bad quality.

Science would counter by saying our observation in that case is merely subjective opinion and not scientifically valid. It is not a "truth" by absolute and objective means. But what if it was? If they could predict the elements of success, why wouldn't a scientist want to be able to predict whether a chain or store or product would be massively successful? Wouldn't this be a worthwhile area of study? I think they would it's just that science, as currently practiced, is hamstrung and fundamentally unable to do so.

Why? B/c it too often conflates quality with measurable innate-ness, whereas, in truth, quality exists between things, it exists in the relationships between things.

To date in history we have pursued measurable innate truths to their logical ends where deeper discoveries require orders of magnitude more time and energy to reap infinitesimally smaller benefits. At a micro level we are bumping against quantum behaviors and the inconsistencies that creates. At a macro level our current technology can only see so far into the distance.

Personally, I think we are at a critical juncture in history where, if we accept a more holistic framework for reality, one that includes both the subjective and objective within the same model and with the same level of validity, it will open up a whole new world of discoveries and truths for us to explore.

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .