Having promised myself not to get involved in any instant coffee student philosophy discussions, I'm going to have to tread quite carefully around this. But Robin poked me on the subject and it does merit some thought. This still officially counts as an interlude ... I am loathe to attempt to formally define or question the concept of 'truth'. And I have previously stated that for current purposes, I'm assuming science is not a list of absolute facts (or truths).
The earth is flat.
What do we mean when we say something is true or false? I mentioned the flat earth model earlier so let us start with that. The statement "The earth is flat" is definitely false. (False is much easier to spot than true.) Here is a collection of photos of the earth taken from space. Conspiracy theorists may wish to dispute their veracity ... but not with me. Thanks all the same.
A flat earth model is very useful if you are building a house, but architects using the flat earth model know that the earth isn't actually flat. I hope.
What about the statement "The earth is round"? If you were pedantic to the point of being antisocial, you might argue that strictly speaking the earth is an oblate spheroid and not perfectly round at all. Is the statement still true? It's fairly accurate. It's probably the best single syllable description of the shape of the earth. It's not as if 'round' is a strictly-defined geometric term in any case. Biscuits are round. People's heads are round. Round-ish.
So in common usage 'true' and 'false' are not absolute black and white concepts.
The earth is 7000 years old.
Some biblical literalists use the creation stories in the Old Testament to assert that the earth is only six or seven thousand years old. Some will even couch this assertion in scientific terms and claim it can be supported by a legitimate use of the scientific method. A quick google will bring up a number of such papers.
This assertion is false. Uncontroversially and incontrovertibly false. This model of the earth's history is contradicted by a truly massive weight of evidence. It is as easily falsifiable as the suggestion that the earth is flat.
Note that this does not instantly disqualify it from being a scientific model. As I said earlier:
"Scientific models and theories are not 'true' or 'false'. They are 'useful' or 'not useful' in particular contexts."
and
"Science is a set of well-defined models that can be tested against observation."
If you can find a context where the assumption that the earth is 7000 years old has an application, and you can express it in terms that allow it to be tested against observation in that context, then it will become a useful scientific model.
I will buy an iPod for the first person who can give me an example of a context where this assertion is useful and a method for testing it. Any size and colour you like. The answers "to prove the literal truth of Genesis" and "by reading the word of God" will receive a mystery booby prize.
So ... being demonstrably false does not stop something from being science; failing to satisfy the simple criteria of a scientific model and not applying the scientific method stops something from being science.
I have no idea if this satisfies Robin's qualms about "true" versus "useful". I'm sure he'll tell me.
Showing posts with label true and false. Show all posts
Showing posts with label true and false. Show all posts
Friday, 25 January 2008
Monday, 7 January 2008
Definition 2: Science
A brief aside whinge about Science.
'Science' is a term that is much abused and misunderstood. Perhaps not as much as the term 'God' ... but it's not been around for quite as long and is doing its best to catch up.
If I mention science anywhere in this blog then this is what I mean. You may not agree with this definition, but it's the one I'm using and I'm the one doing the typing.
That's it. Well almost. I'll expand a bit after this partial list of things that science is not:
What do I mean by 'a model'? Tricky one. Something that looks a bit like the thing it is modelling but is simpler and easier to work with. A globe is a model of the earth. Gravity is a model of how things fall down. Euclidian geometry is a model of how shapes fit together. The important thing is that models are not the same as the things they are modelling. Physics is not 'how the universe works', physics is 'a model of how the universe works'.
What do I mean by 'well-defined'. Trickier one. Something is well-defined if somebody other than its inventor can understand it according to some shared language and rules. This is the difficult bit of science that can lead to impenetrable jargon and funny-looking equations. You can describe a scientific model in plain English; the short-hand simply allows you to state it in a lot fewer pages.
What do I mean 'can be tested against observation'. This bit should be self-explanatory. A scientific model should tell you 'stuff' about the 'thing' that it models. You need to be able to compare the stuff the model tells you with the stuff that you can observe directly from the thing itself. If you can't compare one with the other then it is not a very good model, is it?
THIS NEXT BIT IS IMPORTANT.
Scientific models and theories are not 'true' or 'false'. They are 'useful' or 'not useful' in particular contexts. That is, they either tell you something useful and accurate about the thing they are modelling, or they don't.
Example 1: the world is flat
Hopefully everybody reading this thinks the above statement is 'false'. However, it is the model of the world used by the architects and engineers designing almost every building on the planet. Because it is useful in that context.
Example 2: Newtonian or Classical mechanics
This is an 'old' model of how objects move and interact. It is perfectly useful in almost all human-scale contexts. You can put a man on the moon with Newtonian mechanics. However, there are things that we observe about the universe that do not look like the Newtonian model. So we have to think up a new model that matches these observations. This does not make the model FALSE, it simply means it has a finite utility. It doesn't model everything.
The job of science is to test the models it has already invented until they break. And then invent some new ones to pick up the pieces.
When people dismiss science with the argument that 'science often gets things wrong' they are only partly correct. Science always gets things wrong. That's how science works.
'Science' is a term that is much abused and misunderstood. Perhaps not as much as the term 'God' ... but it's not been around for quite as long and is doing its best to catch up.
If I mention science anywhere in this blog then this is what I mean. You may not agree with this definition, but it's the one I'm using and I'm the one doing the typing.
- Science is a set of well-defined models that can be tested against observation.
That's it. Well almost. I'll expand a bit after this partial list of things that science is not:
- a list of absolute facts
- a system of belief
- the stuff that makes the universe work
- a conspiracy of wicked men trying to kill God
What do I mean by 'a model'? Tricky one. Something that looks a bit like the thing it is modelling but is simpler and easier to work with. A globe is a model of the earth. Gravity is a model of how things fall down. Euclidian geometry is a model of how shapes fit together. The important thing is that models are not the same as the things they are modelling. Physics is not 'how the universe works', physics is 'a model of how the universe works'.
What do I mean by 'well-defined'. Trickier one. Something is well-defined if somebody other than its inventor can understand it according to some shared language and rules. This is the difficult bit of science that can lead to impenetrable jargon and funny-looking equations. You can describe a scientific model in plain English; the short-hand simply allows you to state it in a lot fewer pages.
What do I mean 'can be tested against observation'. This bit should be self-explanatory. A scientific model should tell you 'stuff' about the 'thing' that it models. You need to be able to compare the stuff the model tells you with the stuff that you can observe directly from the thing itself. If you can't compare one with the other then it is not a very good model, is it?
THIS NEXT BIT IS IMPORTANT.
Scientific models and theories are not 'true' or 'false'. They are 'useful' or 'not useful' in particular contexts. That is, they either tell you something useful and accurate about the thing they are modelling, or they don't.
Example 1: the world is flat
Hopefully everybody reading this thinks the above statement is 'false'. However, it is the model of the world used by the architects and engineers designing almost every building on the planet. Because it is useful in that context.
Example 2: Newtonian or Classical mechanics
This is an 'old' model of how objects move and interact. It is perfectly useful in almost all human-scale contexts. You can put a man on the moon with Newtonian mechanics. However, there are things that we observe about the universe that do not look like the Newtonian model. So we have to think up a new model that matches these observations. This does not make the model FALSE, it simply means it has a finite utility. It doesn't model everything.
The job of science is to test the models it has already invented until they break. And then invent some new ones to pick up the pieces.
When people dismiss science with the argument that 'science often gets things wrong' they are only partly correct. Science always gets things wrong. That's how science works.
Labels:
euclid,
newton,
science,
scientific model,
true and false
Subscribe to:
Posts (Atom)
