Saturday, 28 February 2015

Do scientific explanations actually explain?

To provide a scientific explanation of a given phenomenon is to show how it is ultimately derived from physical laws. Sometimes physical laws governing some aspect of physical reality can be shown to be derived from more fundamental physical laws.  But these more fundamental physical laws appear to be simply a brute fact.  In other words certain physical laws appear to govern the Universe and no one has any idea why such physical laws assume the form they do.  Moreover, by necessity, an answer to this question could not come under the province of science since science assumes such physical laws at the outset.

But then scientific explanations are not in fact explanations in the fullest sense of this word.  It is of little avail to say we can explain the existence of X in terms of Y, if Y itself has no explanation and is simply a brute fact. This is why, contrary to what most people believe, science doesn't strictly speaking provide explanations, but rather mere descriptions. Yes we can explain how a phenomenon is deduced or derived from physical laws, but this fails to provide any ultimate explanation if we don't know why physical laws assume the form they do. Indeed since physical laws are simply a brute fact without any explanation, then in providing a scientific explanation we are only ever engaging in discerning specific patterns (some phenomenon) deduced from more general patterns (physical laws).

28 Dec 2015 Update:

A rather excellent article on the same topic by
Dr. Bernardo Kastrup which I'm in full agreement with.  He employs a splendid analogy which is worth quoting:

[O]ne needs to know nothing about computer architecture or software in order to play a computer game well and even win; just watch a five-year-old kid. Playing a computer game only requires an ability to understand and predict how the elements of the game behave relative to one another: if your character shoots that spot, it scores points; if your character touches that wall, it dies; etc. It requires no understanding whatsoever of the underlying machine and code upon which the game runs. You can be a champion player without having a clue about Central Processing Units (CPU), Random-Access Memories (RAM), Universal Serial Buses (USB), or any of the esoteric computer engineering that makes the game possible. All this engineering transcends the “reality” accessible empirically from within the game. Yet, the scientific method limits itself to what is empirically and ordinarily observed from within the “game” of reality. Scientific modelling requires little or no understanding of the underlying nature of reality in exactly the same way that a gamer needs little or no understanding of the computer’s underlying architecture in order to win the game. It only requires an understanding of how the elements of the “game,” accessed empirically from within the “game” itself, unfold relative to one another.
On the other hand, to infer things about what underlies the “game” – in other words, to construct a metaphysics about the fundamental nature of reality – demands more than the empirical methods of science. Indeed, it demands a kind of disciplined introspection that critically assesses not only the elements observed, but also the observer, the process of observation, and the interplay between the three in a holistic manner; an introspection that, as such, seeks to see through the “game.”


  1. Hi Ian: Great post. Freda Matthews has some wonderful articles on the limitations of "laws of physics" as an explanation. Steven Weinberg, in his 2003 New Yorker essay "Does Science Explain Everything? Anything" comes to the same conclusion (with great reluctance) as you.

    I describe this in my "Shaving Science With Ockham's Razor" over at If you ever get a chance to look at it, please let me know.

  2. you make the argument that to play a computergame you dont need to know how the engine works. This much is true however if you're going to go around to tell people how the engine works in fine detail. You can only do that if you actually study the code that is used.

    Beleaving some book over science is like saying.

    hey i read on gamespot that this engine is a miracles.
    And then go talk to someone who ports the game to linux and tell him your knowlage about the engine is illegit becouse gamespot says the engine is a miracle.

    Ofc i dont acpect you to undestand what i'm saying becouse you are probably no an avid gamer.

    1. Hi Mr Unknown. I was quoting someone else -- a certain Dr. Bernardo Kastrup -- in the analogy with computer games. I took the liberty of contacting him and bringing to his attention your devastating critique. He said:

      "Oh my God, he wiped the floor with me... ;-)"

      BTW I do play space invaders, pacman and galaxian, at least when I've had a few beers. Does that constitute an avid gamer? Got to confess, I'm really excited about Virtual Reality!