# Physics the Why vs. How question?

I've been told that a physicist never asks why and must always ask how. Is this true? Have all discoveries in physics been due to asking how? or have some been discovered by asking why?

-
"Why hydrogen atom is stable and has discrete spectrum" may give a hint of quantum mechanics? –  user26143 Dec 14 '13 at 2:49
I'd rather hear a hypothesis on how that could happen. –  Jitter Dec 14 '13 at 6:13
First of all, you should give a precise definition of "why" and "how". In many cases they are equivalent. And you could use other words, such as "what is the mechanism behind?". In fact, I think you don't need to use questions at all (you could say something like "I want to know the reason of..."). –  jinawee Dec 14 '13 at 17:24
I took the liberty of asking appropriate mirrors to this question at philosophy.se and english.se. Hopefully, the different perspectives of these stackexchanges will give us some additional insight into this question. –  Geoffrey Dec 14 '13 at 19:51
More on Why vs. How: physics.stackexchange.com/a/54918/2451 –  Qmechanic Dec 19 '13 at 13:20

English is pretty imprecise and asking questions of the form "Why does ..." feels natural even when "How does ..." would suffice. "Why" has philosophical overtones because it goes beyond just "How". Physics is a science and the scientific method is tried-and-true method for gaining knowledge about how things work.

I'd suggest that in any place you have a question "Why does ..." try to substitute it with "How does ..." instead. Any portion of "Why does ..." not answered by "How does ..." is the philosophical / non-scientific portion of the question.

For example "Why does 1 + 2 equal 3?" can be substituted with "How does 1 + 2 equal 3?" and answered adequately. Depending on how deep and technical you want to get, the answer may delve into the definition of addition, various axioms, peano arithmetic, and so on. If you still have a question about "why" after all of that, your question is of a philosophical nature.

Physics is firmly rooted in the scientific method, mathematics, and experiment. It simply can't answer the philosophical portions of questions.

-
Do you feel why captures better the question of understanding the mechanism by which like charges repel and opposite charges attract? (Why do like charges repel? vs How do like charges repel?) Part of me feels this way. I expect logical deduction from theory for a why question (I want the word because in the answer), whereas I don't have that expectation from a how. –  BMS Dec 14 '13 at 1:40
If I asked, why does one plus two equal three then the answer should be because that it how logic works? –  Jitter Dec 14 '13 at 1:42
You're right that "how" often feels mechanical in its description of things while "why" feels more like a logical, deductive process. I think feeling this way is a bias we get from how we use English naturally versus scientifically. How does include all of the logical deduction and reasoning too. –  Brandon Enright Dec 14 '13 at 1:42
@Jitter that would be a very poor answer. "How 1 + 2 equals 3" does include all of the reasoning behind it. "How" goes into all of the depth you want from "why" without any of the philosophical baggage. See my above comment. How doesn't have to be a list of facts. It can (and should) include the deductive reasoning too. –  Brandon Enright Dec 14 '13 at 1:45
bms, l was just thinking about that the other day. If you throw virtual balls at each other you repel, that's the easy one. If your virtual balls follow a curved path and follow a field line then you get attraction. –  Jitter Dec 14 '13 at 1:51

I agree with the others that physics is fundamentally a "hard science" which is (or should be) concerned exclusively with physical, empirical facts, but I have to disagree with the assertion that "How?" is a scientific question while "Why?" is not. Moreover, philosophy is central to scientific research despite not being itself the end goal. This discussion with all of its deep linguistic analyses bears witness to the importance if not the usefulness of not only understanding nature but also understanding how we understand nature.

Plenty of sound scientific questions have begun with "why." Faraday wondered why current deflects a compass needle, and why the galvanometer attached to a circuit jumped whenever he closed the switch on another circuit. Many scientists of the early twentieth century asked why beta decay appeared to violate otherwise sacrosanct conservation laws. Newton asked why Kepler's Laws were so accurate. None of these questions can be adequately rephrased as "how" questions because they all arise precisely from a deep and thorough understanding of "how," an understanding left frustrated by inadequate explanation.

For this reason, these sorts of questions - the questions of curiosity that challenge the limits of our theories and guess at what might lie beyond - are exactly those which allow science to grow and flourish. Remember: science is only as good as the people who practice it, and no theory is truth, every theory (no matter how accurate and well-established) is somebody's attempt to make reality more palatable.

"How" is necessary and important, don't get me wrong, but so is "why." As Thomas Kuhn might put it: "How" is status quo science, journeyman science, science "as usual"; "why" is fringe science, confrontational science, paradigm shift.

The truth is that it is disingenuous to sacrifice "why" on the altar of "how." Forcing our language into some tightly defined pen where only "how" is allowed is simply a false dichotomy born of a mistaken appreciation of natural language.

And just to satisfy all of the empiricists and positivists out there, allow me to present you with test case number 1 for "how" gone wrong: thermoelectric theory. Thermoelectric theory in the mid twentieth century was just plain wrong. It was built on a flawed phenomenological model that was not adequately challenged at the time, and as a result, it was not until pretty recently that anyone thought that it was even possible to produce a material with a Seebeck Coefficient greater than one. That model was built on "how." Like all phenomenological models, it did its best to fit experimental data simply and elegantly, and many people failed to miss the point that it fit experiments so well because it was made to do that. It wasn't predictive, but it was accurate in the only cases that were ever tested. No one complained about the "why" since the "how" seemed so right, but it wasn't.

Without "why" to challenge "how," we're nothing but theoretical engineers or practical mathematicians. Physics - true physics - is fathered by "how" but born of "why." Speaking for myself and many others, I love and study the how, but only because I am intoxicated by the pursuit of why. "How" may be the wheels on the road, but "why" is the engine of the car. We need them both to move forward.

Edit: After the useful conversations at philosophy.se and english.se, I feel that it's worth adding one more idea here. (Or perhaps just saying an old idea another way.)

"Why" questions challenge theories or speculate about possibilities. They allow us to address the most difficult issues of science which develop when we run up against the boundaries of our knowledge. Furthermore, they are useful when the problem at hand is characterized by hidden information or unknown parameters which are strongly affecting the experiment. For instance, in the beta decay example above, it was not possible to ask "How do neutrinos affect beta decay?" because the idea of neutrinos had not been invented yet; in fact, "Neutrinos." is precisely the answer to the question "Why does beta decay seem to violate conservation of momentum?"

But, of course, neutrinos could only be postulated because so many scientists asked the questions "How does beta decay behave under these conditions?" or "How does it behave under those conditions?" By accruing data (which are the answers to "how" questions), they were able to identify discrepancies between observation and expectation which required new theories and creative thinking to adequately explain.

"How" questions generally are more tightly focused and lend themselves to being phrased as implicit hypotheses which can be tested, but they necessarily live within the context of a theory and, therefore, presuppose some fore-knowledge. "Why" questions, on the other hand, generally strike at the heart of a scientific issue by identifying defects or peculiarities in a theory which might lead to new science. By doing so, "why" questions need no theory and may pursue an explanation of observation without reference to a pre-established groundwork.

I think that "how" questions are extremely important in the actual practice and study of science, but "why" questions embody the ever-striving, almost combative quality that peer-reviewed science takes on when theories compete with one another for acceptance and dominance.

-
Your answer is written like an evangelist's rhetorical essay and love letter to "why" over "how". Science does need a certain amount of evangelizing to maintain funding and respect in a very un-scientific world but I think within the realm of science it's best just to stick to the facts and methods for learning and testing those facts. –  Brandon Enright Dec 14 '13 at 3:37
All we have to observe is reality; reality is the given. The question of "why is reality this rather than that" presupposes a cause outside of reality and is, thus, non-physical. –  Alfred Centauri Dec 14 '13 at 3:37
@BrandonEnright: I agree: the facts are the only important thing when it comes to science. I don't understand how the pursuit of deep understanding of nature is contrary to this precept. Moreover, my central point is that learning and testing facts must be done by humans who think in terms of "why" and use "why" as a tool to pry open the oyster of reality. "Why" should be embraced as a part of the scientific dialectic, not eschewed like some airy-fairy activity enjoyed by effeminate, pipe-smoking yahoos. –  Geoffrey Dec 14 '13 at 3:47
@AlfredCentauri: I agree that reality is all we have to observe, but the question of why reality is this way rather than that has many uses (which is a fact demonstrated innumerable times over the course of history). It is not non-physical; there is nothing more physical than the pursuit of a detailed and accurate theory of reality. "Why" is the way that we plumb the depths of reality and bring back to the surface usable insights about the universe. –  Geoffrey Dec 14 '13 at 3:51
@Geoffrey, you seem to be unclear on the distinction between "why" and "how". The pursuit of a detailed and accurate model of reality comes from asking how, "how does the universe work?". Consider the vastly different question why does the universe exist this way or at all?. Note the vast difference between the questions "how are we here?" and "why are we here?". –  Alfred Centauri Dec 14 '13 at 13:38

In English, asking 'why' is a request for an explanation of, reasons for, or purpose of something.

Physics does provide explanations and some types of reasons, but doesn't concern itself will questions of purpose at all, at least not in any teleological sense.

So physics answers some types of why questions but not others, because in English usage 'why' is overloaded and can be used to ask for many different kinds of things.

-

"Why" is the question for philosophers. "How" is the question for physicists.

The task of a physical scientist is to find a model that adequately describes "how" nature behaves. When more than one model fits, see Occam's Razor.

The question of "why" nature is described by one model or another is best left to the philosopher and, be aware, that path leads to madness.

-
Madness, or else original discovery. –  Geoffrey Dec 14 '13 at 4:16

"Why" is the fundamental scientific question, the question regarding cause and effect. "How" already implies that certain things cause others, and sometimes it seems that cause and effect are not that easy to distinguish. You can't do any science without "why".

Take, for example, the question: "How does CO2 contribute to climate change?"

This is a scientific question only after it has been established that

• the climate changes
• it does this because of green house gases in the athmosphere
-
Why isn't needed to formulate either question: "Is climate changing? If so, how?" and "Does $\mathrm{CO}_2$ contribute to climate change? If so, how?" –  Brandon Enright Dec 17 '13 at 16:37
@BrandonEnright To the first question one could answer "By getting warmer." So, where is the cause? There are plenty of possibilities. If science is finding what causes something observable, then "why" is a must, isn't it. –  Ingo Dec 17 '13 at 16:40
"By getting warmer" is not a good answer. There is nothing special about why that gets good answers while how allows incomplete or poor ones. Everyday English usage does elevate why to get at meaning and how to just mechanical cause-effect answers but these are biases in how we use our language and not part of the actual definitions of how and why. Science is free to use them correctly and in doing so, it avoids the extra non-science baggage that comes with why. –  Brandon Enright Dec 17 '13 at 16:44
@BrandonEnright So, then maybe it's me since I am not a native speaker. In my language (german) "why" asks for the cause, and "how" for a description of the phenomenon. In fact, for the question "Wie ändert sich das Klima?" (how does climate change?) the best appropriate answers would be to describe the observations that have been made, and the second best would be to descibe mechanisms like radiation, green hoeses and so on. But "Was ist die Ursache der Klimaänderung?" (What is the cause of climate change? or simply Why does it change?) is a whole different question. –  Ingo Dec 17 '13 at 16:52

Physics is concerned with abstract, mathematical models that describe the behaviour of particles, space, time etc and that are tested with observations and experiments. If a model is in agreement with experiments, under certain conditions and up to a certain accuracy, then it is accepted (for situations where those conditions hold). But a theory remains an abstract model; it provides a way in which nature can be described, but doesn't say what nature is. So one should always be cautious about interpreting theories and taking those interpretations too far.

Richard Feynman illustrated this in one of his lectures (his argument starts around 5:30 in the first clip and end around 7:30 in the second clip):

The Relation of Mathematics and Physics - Part 4

The Relation of Mathematics and Physics - Part 5

In the clips, Feynman shows that Newtonian gravity can be described in 3 different ways: action at a distance, a potential field, and the principle of least action. Conceptually, these 3 descriptions are very different. However, mathematically, they give exactly the same results. This means that, as far as physics is concerned, all three are equally valid and one has no reason to prefer one above the others. There is no unique "right" answer.

In fact, as Feynman argued, it is a bonus to have more than one description, because they enable us to look at new problems from multiple angles, and they offer us more routes and tools to find new theories.

-

anna v wrote in her answer:

The only answer to "why" is this "because", data says so.

I'd like to provide another example. Consider the question "How does matter move through space". In trying to answer this, Aristotelian arguments could satisfy someone of the answer. Prodding further, Aristotelian arguments don't hold up and we need to use Newtonian arguments. Even further, one could do a double slit experiment and show that, in fact, a Newtonian particle description is inadequate and a quantum mechanical description is needed. In each of these cases there already has to be a framework where we're satisfied with an answer. I agree that "$\dot{x}=\text{constant}$" could be a description of "how" the particle moves, but that's assuming we're working with a Newtonian viewpoint. If we don't allow ourselves to use a Newtonian viewpoint, then that's not how the particle moves at all!

So, if answers to "why" are, "because the data says so" or "because that's the idealization we're considering", how are answers to "how" different? (and why?!)

In the related philosophy.se post, ChristopherE writes:

Your other question is "Are 'why' questions" (broadly understood) useful in or applicable to the study of science?" Certainly, yes, and there needn't be anything revolutionary about them.

I don't think this "why" vs "how" business gets to the heart of the matter, and of course the linguistic nature of "why" and "how" obscures things further. Maybe it would be better rephrased as "physics doesn't prove things in the mathematical sense". Ask "why" "what" or "how", but don't suppose you get a leg up on nature just because the resulting models are predictive.

disclaimer: I'm a baby physicist

-

Here are my two cents of the euro:

Physics is a science that has a large body of observations, and a limited number of mathematical models/theories that aim to organize and explain those observations and , very important, get validated by predicting the behavior of new observations.

Mathematical theories start with axioms and some tools that develop theorems from those axioms and then various setups can be examined.

For example Euclidean geometry starts with axioms and ends with being able to predict and design complicated geometrical shapes. One can start asking why the sum of angles in a triangle is 180 degrees and one can prove it using the tools. If one goes further up in the why questions one ends up with the axioms. If one changes the axioms then a different geometry is implied. So "why" questions end up on the axioms. One could as well ask "how" one gets 180degrees for the sum of the angles of the triange, and then the "why" goes to "why start with these axioms".

In physics the theories in addition to the mathematical construct has equivalent to axioms, postulates. These have been postulated because of the need for the mathematical model to agree with measurements and data in general. For example the Heisenberg Uncertainty principle. Or the "square of the wavefunction gives the probability of the measurement". In a similar manner as in the above mathematical example, all the "why" questions in physics are really answered as "how" one goes from the axioms and postulates to the specific observational data or predictions. The "why" ends up on the axioms for the mathematics and postulates for physics. And the answer then is "because" these basic assumptions/postulates are necessary to fit our mathematical model to the existing data and give us confidence in predictions for new observations.

The only answer to "why" is this "because", data says so.

Edition after comment by Ingo :

Once one has a theory, and physics is really as I stated in the first paragraph a mathematical theory that organizes known data in order to be able to predict future unmeasured ones, the "how" question gives the causal path in our understanding of how the final data/observations happened and how the predicted ones will appear. Why questions address the existential state. When we have no theory and have an observation we start with "Why", because the observation exists. When a theory forms then it is the causal path that is sought and "why" goes up the mathematical ladder by "how mathematically this happens" transferring the existential question to the axioms and postulates.

-
What if "why" is used in the sense of "What is the cause of this phenomenon?" ~ "Why is this happening?" –  Ingo Dec 17 '13 at 18:56
@Ingo We can only really show how from the postulates and the axioms of our model for that particular physical phenomenon we can explain this phenomenon. Then one goes into "why these axioms and postulates", which are only answered with "because that is what fits the data". –  anna v Dec 17 '13 at 19:36
I am only interested in the first "why" for now. And I do not understand what axioms we need to experience that often when B happens, it was preceded by A. Then to formulate the hypothesis that A somehow (and this is where the "How" comes into play - I am not saying it is not important, mind you) causes B. And then we do what scientists do - carry out controlled experiments, do some statistics maybe, and finally distill a theory. But the "why happened B?" is at the beginning. –  Ingo Dec 17 '13 at 22:08
@Ingo Sure, that is the process when then we gather data and are on the way of building up a mathematical theory. after a while we say that B happened because it was preceded by A. When we discover that A is always preceded by C and C is always preceded by D etc we are on the way to a mathematical theory which has formed our way of describing data in physics. Then we see that we know how B happens, it is because of D, and if we want B to happen we create the D situation because we have the causal path of how each causes the other. How gives the causal path which is the only thing we can do. –  anna v Dec 18 '13 at 4:10
continued: We have taken to the "Why" to "why D" –  anna v Dec 18 '13 at 4:11