Do the words ‘belief’ and ‘faith’ belong in science?

The words faith and belief obviously have a natural home in religious discussions. Should scientists avoid using such words, as in statements like ‘”I believe in the theory of evolution” or “I have faith in the law of gravity”, since that seems to put them on a par with “I have faith/believe in god” and enables religionists to claim that scientific theories are similar to religious beliefs? In a recent comments section, a recurring suggestion came up that in order to avoid this misapprehension, we should avoid use of the words belief and faith altogether in scientific discussions.

I disagree with this suggestion for two reasons, one practical and the other somewhat philosophical. The practical reason is that language use tends to be anarchic and is impossible to police. People use language any way they want and indeed that is its source of much of its vibrancy, with new words and usages coming into being all the time as people break existing rules. Witness how hard it is to prevent the new usage of words and phrases that actually have the opposite of their original meanings. The word ‘literal’ is now sometimes used to mean ‘figurative’, the word ‘finite’ is used to mean non-zero or not infinitesimally small when it technically means ‘not infinite’, and the phrase ‘could care less’ is increasingly used as equivalent to ‘couldn’t care less’. There are those who fight valiantly against what they see as these signs of the abuse and corruption of language but they seem to be losing.

Hence asking anybody, including the scientific community, to refrain from using perfectly functional words such as faith and belief in scientific discourse is an exercise in futility, especially since such usage is nowhere near as egregious an abuse of meaning as the examples I gave above.

This brings me to the philosophical objection to trying to restrict their use, which is that the words belief and faith have multiple meanings depending on the context and have their place in science as well. It would be a mistake to concede perfectly functional words to religious people.

We know that it is impossible to prove a scientific proposition to be true. At best we have a preponderance of evidence in favor of it. If I release a rock in the air, it will fall to the ground and I ascribe that behavior to the law of gravity. If I repeat the drop a hundred times, the rock will fall to the ground each time. But that still does not prove that it will do so the very next time I try it. (This is the well-known problem of induction.) But I have every reason to believe that it will do so the next time and this is what is meant by saying that I have faith in the laws of gravity. In fact, I did not repeat the dropping exercise one hundred times but just said it. I did not drop it even once, so sure was I of the outcome. Readers would have thought I was crazy if I had told them that had carried through this exercise just so that I had evidence to back up my statement. When I quote this example to religious people, they too do not question the premise that the same result will obtain however many times it is repeated. And yet, we cannot say that the law of gravity is 100 % proven. We only believe that it is true and have faith that the rock will fall again.

In that sense, I am using the words faith and belief in the same sense that religious people use it, as saying that I have total confidence in its correctness even in the absence of absolute certainty. And we need to have that sense of certainty. If I had doubts about the established laws of science continuing to hold in all situations, I would be unable to function, just like those who have doubts about the laws of fluid mechanics avoid flying in airplanes. I need to be certain that when I open the faucet, the water will flow down into the sink and not spray all over the kitchen and ceiling, so that I can do it without thinking. Life would be a nightmare without our faith in the infallibility of scientific laws.

But there is a big difference between scientific faith and religious faith. Scientific faith is what enables us to bridge the small gap between near certainty (based on a preponderance of reliable and predictable empirical evidence) and certainty. Religious faith is what enables people to bridge the huge gap between unbelief (due to the lack of reliable and predictable evidence) and certainty. In the case of science, we try to make the gap to be crossed by faith as small as possible, though we can never eliminate it. If the gap starts getting bigger due to the appearance of anomalous evidence, that is a cause for concern. It is not a virtue to cling to certainty when the gap increases.

Contrast this with religious faith. There it is considered a virtue to hold on to certainty, especially when the gap is large and getting larger. Those in the religious world who believe in a god despite the lack of evidence in favor are praised for their faith, whereas in the scientific world those who believe in some theory despite the lack of evidence are viewed with concern. When a tragedy strikes innocent people, which should be considered clear evidence against a good god, people are praised for continuing to believe, even though the gap has now got larger.

What we should realize is that people learn to ascribe meaning to words not only on the basis of their strict dictionary meanings but also on the context. As a result of repeated usage, there is now usually no problem in understanding what people mean when they use ‘literally’ or ‘finite’ or ‘could care less’ in everyday conversation and writing. We should aim to achieve the same level of contextual understanding of the words faith and belief too. So rather than avoid the use of those words in science, we should use them freely and when people question us about it, take that as an opportunity to explain the difference between scientific faith and religious faith, between scientific belief and religious belief.

Hence when I give talks to students about how being scientific is incompatible with believing in any form of the supernatural, there are always some clever students in the audience who pick up on my use of the words faith and belief and argue that this implies that even I implicitly think that religion and science are similar knowledge structures. I use that as an opportunity to discuss the difference and I think that those discussions really help to advance their understanding of the nature of the scientific enterprise. I think the outcome is better than if I had avoided those words altogether.


  1. says

    In that sense, I am using the words faith and belief in the same sense that religious people use it, as saying that I have total confidence in its correctness even in the absence of absolute certainty.

    No, you’re not, and you should stop letting persons of faith pretend you are, because that allows them room for all sorts of blatant dishonesty and obfuscation. When you have confidence in a material event, like gravity always working the same way, that confidence is based on a) a preponderance of evidence that gravity works an certain way, and b) absolutely zero evidence hinting that it might change tomorrow or next year. When a believer expresses confidence in a supernatural belief, he does so based on little ro no evidence at all, and probably ignoring lots of evidence that refutes his belief.

    It would be a mistake to concede perfectly functional words to religious people.

    They’ve pretty much commandeered those words already, to the point where most of the people who hear them hear the religious meaning, unless they’re couched in terms that tend to negate it. “I believe in gravity” inevitably sounds like a statement of ungrounded faith; you’d have to say something like “I have reason to believe…” to get around that misunderstanding.

    If I repeat the drop a hundred times, the rock will fall to the ground each time. But that still does not prove that it will do so the very next time I try it. (This is the well-known problem of induction.)

    Actually, for all practical purposes it kinda does, at least until you find new evidence or reasoning indicating the possibility that the behavior of objects might change in the foreseeable future.

  2. left0ver1under says

    The words “belief” and “faith” belong only in the descriptions of emotions and actions.

    “I believe students should study science.”

    “I broke faith with religious nonsense.”

    Using the vocabulary of the religious only aids them in forwarding their ignorance.

    Scientific words should be used to describe scientific matters:

    “I have confidence in evolution.”

    “The Big Bang is a certainty.”

  3. Makoto says

    I don’t like using the word ‘faith’ to describe my ‘trust’ in science. I have reasons for trust, while faith (of the religious flavor) specifically requires a lack of reason – there is no proof of god, gods, reincarnation, the soul, afterlife, etc, nor even testable hypotheses about how those could exist. I can trust that the explanation of how and light bends is correct, given the evidence and theories of why it does what it does.

    Gravity is a really interesting example to me. We know that it works by observation, but it’s very difficult to experiment upon. We know enough about it to fling an object from the surface of our planet and get it to land with astonishing accuracy on another, but we can’t yet build a device to increase or decrease the effect of gravity in an area because we can’t alter it without altering other physical characteristics about the objects involved.

    Contrast that with something like evolution, where we have both observational evidence and experimental evidence showing details on how it works, along with the theories to show why it does. We’ve controlled it to an extent, and have been for a long time, by replacing the natural selection portion with artificial selection to bring forth desired traits in organisms.

  4. sunny says

    You are preaching to the choir here. While you and other scientists may be using these words in a particular sense, the religious are not. In fact, your faith or your belief is contingent on the evidence: you will change your faith or belief if they shown to be inconsistent with the evidence. This is hardly the case with religious faith or religious belief. In fact, it is the very opposite.

  5. AsqJames says

    Funny the connections/parallels our minds make – Mano’s musings brought to mind last Thursday’s In Our Time on Bertrand Russell ( Russell wouldn’t accept as axiomatic (i.e., he wouldn’t take it on faith or believe without evidence) that 1+1=2 and set about proving it logically. Which in turn brings to mind the XKCD “Purity” cartoon ( I’m thinking another way to measure the “purity” of a scientific field is to quantify the gap between the available evidence for the field’s accepted theories, and what would be needed to satisfy someone like Russell that the theory had been proved.

    …and now I’m wondering how much white space you’d need on the left of the sociologist for a theologian to be accurately positioned?

  6. says

    I think those words have been ruined by their association with ignorant fideism. So I prefer to use evidence-based language, stealing a riff from Sextus Empiricus, “It would appear to me now that…” or “It seems likely that…” or “I accept evidence that shows…”

  7. jamessweet says

    As to the philosophical issue: I agree with makoto. I have no problem with using the word “belief” here, but I think it’s a mistake to use “faith” as synonymous with “trust”. While it is true that the word “faith” can be used that way, there are perfectly serviceable words (like “trust”) that mean the same thing and don’t have the ambiguity that it could mean something like religious faith.

    And while I think that “belief” is in many ways meant in the same way in both contexts, the type of “faith” we are talking about could not be more different. It is not merely a matter of degree, as you suggest when you contrast bridging the gap from “near certainty to certainty” vs. “unbelief to certainty” — “trust” as I will call it is a sort of earned belief-without-evidence. It is belief based on indirect/inductive evidence, as opposed to a belief based on no evidence at all. It is true that I do not have any direct evidence that gravity will work the same way tomorrow, but the theory of gravity has established enough trust that it has sort of earned it on an inductive basis.

    Perhaps a different example would work. Say we have a particular theoretical physicist who is establishing a new cosmological model. The model has yet to make any testable predictions, but it fits existing data so well that she is rationally justified in being “nearly certain” that it is true. If she were nevertheless “certain”, even though only near certainty were justified, we might say that she had something like religious faith, albeit to a different degree. The model has not yet made the kind of predictions it would need to earn that “trust” that we put in gravity, so we might call that “faith”, but only a little bit of faith. I think that is a thing which is different in kind from the kind of “trust” we place in the theory of gravity.

    Now, I do think that the Problem of Induction can only be solved through a tiny leap of “faith”, a faith that is the same in kind as religious faith (though to a much different degree). But once we have “faith” in induction, particular theories can rely on it for justification and IMO only be said to require “trust”.

    As to the practical issue: Yeah, people will use “faith” to be synonymous to what I am calling “trust” here, and there’s not a lot we can do about it. I think it’s a mistake, because “trust” is unambiguous while “faith” is very ambiguous. But the best I can do is a) avoid using “faith” in that sense myself, and b) if somebody talks about it, share my opinion. People will surely continue to use “faith” to mean “trust” whether we like it or not.

  8. mnb0 says

    “I have every reason to believe”
    It’s really easy to rephrase this: “I have every reason to assume ….”
    Assumption refers to empirical confirmations. Belief refers to faith instead.
    I don’t have faith in the laws of gravity; I know they correctly describe countless data.
    While I agree with your remarks on the nature of daily language it’s also true that every single individual can influence it. If we materialists want to distinguish ourselves from believers we should have our own terminology.
    The Problem of Induction perfectly shows why I prefer assumption. I assume that the rock will fall because of a) all the experiences of the past, ie extrapolation; and b) a solid theory that describes those experiences plus makes an unambiguous prediction. Assumption still leaves room for some doubt, which is necessary if we realize that some 25 years ago two Germans won the Nobel Price by refuting a theory that also won a Nobel Price – I’m talking about superconduction and BCS.
    It feels misplaced to me to say that I believe in BCS and have (had) faith in its predictions.

  9. Glenn says

    Belief in the supernatural requires faith in the absence of interrogatively derived evidence.

    Belief in the natural (science) arises from the absence of interrogatively derived evidence to the contrary.

  10. says

    I am fine with using the word “belief,” but to use “faith” is a mistake. Faith is the word the religious like to hide behind when you’ve demonstrated their claims to be false. It’s not just bridging a large gap, as you suggest; it is often used to get around logical contradictions.
    Another issue I have is where you say, “What we should realize is that people learn to ascribe meaning to words not only on the basis of their strict dictionary meanings but also on the context.” I don’t find this to necessarily be true. I have seen the word “faith” used numerous times in equivocation fallacy precisely because context was not considered. I find that people frequently ascribe meaning to the word “faith” based on the contexts they use the word, not necessarily in the context that the person speaking the word intends. In other words, if you say you have faith in gravity, you may very well have a different context in your mind than your audience. This is a problem, especially given the degree of difference between how you use the word and how the religious use it.

  11. Irreverend Bastard says

    But there is a big difference between scientific faith and religious faith.

    Exactly. Just like the problem of scientific theory. When a scientist uses the word “teory”, the religious hears “drunken hypothesis, completely unsupported by any evidence”.

    I don’t believe in the theory of evolution, I understand it. Likewise, I don’t have faith in the law of gravity, it’s an indisputable fact of life.

    Let’s not confuse the religious by using “reserved words” in unexpected ways. Language should help communication, not hinder it.

  12. alanuk says

    Truth and belief are uncomfortable words in scholarship. It is possible to define as true only those things that can be proved by certain agreed criteria. In general, science does not believe in truth or, more precisely, science does not believe in belief. Understanding is understood as the best fit to the data under the current limits (both instrumental and philosophical) of observation. If science fetishized truth, it would be religion, which it is not. However, it is clear that under the conditions that Thomas Kuhn designated as ” normal science” (as opposed to the intellectual ferment of paradigm shifts) most scholars are involved in supporting what is, in effect, a religion. Their best guesses become fossilized as a status quo, and the status quo becomes an item of faith. So when a scientist tells you that “the truth is . . .”, it is time to walk away. Better to find a priest.

    Timothy Taylor – Archaeologist, University of Bradford

  13. baal says

    Turns out the purer folks usually don’t quite get the ‘muddy’ part right. When I was a grad student in the life sciences, we’d occasionally see a physicist come over and insist we could get rid of our probabilistic (in biology, pretty much everything is on a probability curve) presentations*.

    *excepting the trivial cases such as an elephant stomps on your head. example: what exposure of dangerous chemical X will kill you? turns out it depends on a number of factors. If you test vs a population, however, you can find something usually called the ‘LD50’ or the amount that will kill half the studied population. You could do big numbers and get a decently small set of error bars but they never really go away.

    I don’t see that scientists should give up use of “faith” and “belief.” The words are too useful, however, scientists (we, i think of myself as one) should be ready to explain what we mean by them. It’s part of being prepared for the xians to attack your usage (and they will, and they do).

  14. Jared A says

    Ok, splitting hairs here, but BCS theory was not refuted. It just turned out there are other mechanisms for superconductivity that hadn’t yet been considered. This isn’t a case like Newtonian physics where it is only true in certain regimes. Most superconductors are accurately and properly described by BCS, and it is still the preferred model for the vast majority of superconductors.

  15. Jared A says

    Hold on. Why is everyone pretending like the only common-parlance usage of faith is in the belief sense? Using it to refer to trust is at least as common. Let’s look at the related form of the word “faithful”. When someone says they are faithful to their spouse, it means that the speaker is not violating the other person’s trust. Likewise when we use it to describe an object or person. When I say “my faithful dog” I could mean two things. I could mean he has irrational beliefs about me–which is true–or I could be using the other meaning of the word, which is that he does not act to violate my trust in him. Generally the latter form is how people use the phrase, so much so that using it in the former sense is engaging in wordplay. The two interpretations actually have inverted meanings and that is because the word has multiple meanings. You pick a meaning based on context, which is how language works. There is no such thing as an utterly logical language, and using euphemisms (Always saying “The preponderance of evidence suggests…” instead of “I believe…”) only kicks the can down the road.

    Apparently I’m in the minority here, but I think it is idiotic to throw away perfectly good word usage because you don’t like how some people use it sometimes.

  16. says

    It’s really easy to rephrase this: “I have every reason to assume ….”

    Actually, the word “assume” has a pretty different meaning, at least in rational enquiry: “assume” means to take something as a given, knowing it’s not proven and may yet prove false, purely for the sake of answering a particular question where there are gaps in our knowledge that have to be filled by something. as in: “What will happen to oil prices? Well, assuming no regime change in Saudi Arabia and no currently unforeseen discoveries of new oil fields, I’d say…”

Leave a Reply

Your email address will not be published. Required fields are marked *