The model and the modeled

[geordief]

I have heard it said that, in science all we have is models that are or are not in accordance with experimental data.

If the former ,they are accepted as  faithful guides or descriptions  until such time as new evidence invalidates them in certain areas.

So models are accepted as  the foundation of scientific understanding.

What about "the modeled"?

 

Can we have a model without there being a "modeled" Is this just semantics ?Does the existence of a model require the existence of a "modeled"?

Is the model (as it is refined) chasing some kind of a mirage of a "modeled" with the two never likely to combine as one or is the "modeled" just  a trick of the language  where the active usage normally predisposes a passive usage?

Can we have models without  a "modeled"?

 

Edited November 7, 2020 by geordief

[MigL]

The 'model' describes the 'modelled' where applicable.
The model is usually mathematical and the modelled is usually 'what we observe' ( reality ? )

In NO case is the model a full description of the modelled; models are, by definition, reduced representations of the modelled, so they have some areas where they are no longer valid representations.

[swansont]

A model has to be modeling something. Otherwise it’s just equations.

[geordief]

46 minutes ago, swansont said:

A model has to be modeling something. Otherwise it’s just equations.

Not just modelling the observations?

 

Describing all the relationships between all the observations and predicting what something as  yet unobserved will turn out to be.

 

Like an expanding volume of observations .

 

Like the tyrant eating his figs every day and thinking they are good for his health until one of his courtiers is told "all tyrants must die" and poisons his dessert 😃

[swansont]

49 minutes ago, geordief said:

Not just modelling the observations?

Modeling the observations is modeling something.

 

[studiot]

2 hours ago, geordief said:

Does the existence of a model require the existence of a "modeled"?

Really depends what you mean.

A Civil Engineer might model the performance of 5 dams he could build to select the one he eventually does build.

A planetary scientist might model the atmousphere of a planet 1.5 times the size of Earth at a specific distance from a specific star, although we know of no such planet.

[geordief]

4 minutes ago, swansont said:

Modeling the observations is modeling something.

 

But something we will only ever experience indirectly and in facets?

1 minute ago, studiot said:

Really depends what you mean.

A Civil Engineer might model the performance of 5 dams he could build to select the one he eventually does build.

A planetary scientist might model the atmousphere of a planet 1.5 times the size of Earth at a specific distance from a specific star, although we know of no such planet.

But an untested model is no model at all .It is like a Trump sharpie model of a hurricane 😃

[ahmet]

it seems this is a meaningless topic to me..

[swansont]

1 hour ago, geordief said:

But something we will only ever experience indirectly and in facets?

I don’t know what this means

1 hour ago, geordief said:

But an untested model is no model at all .It is like a Trump sharpie model of a hurricane 😃

No, an untested model is an untested model.

[Ken Fabian]

Sure, we can model imaginary things, like (studiot) dams that have not been built or stars with planets of varying types and sizes, or global warming with varying emissions pathways that have not and may never happen - but these are built around fundamental understandings of the physical processes, informed and bounded by observable or experimental data.

5 hours ago, geordief said:

Can we have models without  a "modeled"?

Imaginary things or  imaginary physical processes could be modeled as well, in combination with real ones - what might happen if CO2 were not a greenhouse gas or how a Big Bang might propagate in a universe with different physical law for example - or with entirely imaginary physics, without any references to reality as we know it. Not sure what purpose the latter might serve - curiosity, sure, but they may give mathematical or other insights.

[joigus]

Do you mean theoretical models or laboratory (experimental models)? There are both, and I think the range of validity varies depending of the model's faithfulness. For laboratory models I think scaling properties are very important. Theory having some more leeway to construct very simple-minded toy models to better understand the theory.

And I think @Ken Fabian makes a very good point, that sometimes models are built just to understand this very important "what if."

[MigL]

Maybe my understanding of a 'model' is not the same as everyone else ?

We have a 'model' for dams, Studiot.
It comprises the interactions of the stresses and strains, with loads, material strengths, and gravity.
We simply 'plug in' different parameters and conditions to arrive at the five different configurations.

Similarly with climate science, Ken.
We have a model of the atmospheric interactions with solar heating, oceanic heat sinks, and greenhouse gas insulation.
We simply 'plug in' the differing parameters/conditions to arrive at a possible outcome.

The 'model' is the theory of the interactions that govern ( where applicable ) the parameters/conditions we wish to investigate.
For example, the theory of gravity is the model which tells us how different masses, with different speeds will orbit.
While EM theory is the model into which we 'plug in' the charge of a particle and we know which direction it will curve in a particular field.

Or did you mean the scale model of an AVRo Vulcan, which you assemble with plastic cement, paint, and hang in your bedroom.
( yes, I had one of those 🙂 .

[geordief]

2 hours ago, swansont said:

I don’t know what this means

I think I misunderstood you to  be suggesting  there was some "thing" that the model was striving to describe entirely accurately and so ,as it were "merge with"   **  whereas I now think you are just saying that the models are modelling observations (the observations are the "something")

But if you think I correctly understood what you meant by "something" I could try to explain what I meant by "only ever experience indirectly and in facets"

 

**ie there would be no distinction between the model and the modelled(I had a  really fundamental model such as quantum gravity in mind, although I expect that such a model will yield to more fundamental models in the fullness of time)

[studiot]

19 hours ago, geordief said:

 

But an untested model is no model at all .It is like a Trump sharpie model of a hurricane 😃

No one with any sense will make no attempt to test a model.

It may not be possible to match the model against the actual thing being modelled (perhaps like my dam because it does not yet exist) or perhaps because it pertains to conditions we cannot achieve on Earth such as processes with a star.

But, it makes sense to 'test' the model under simpler known conditions and hope that the extrapolation will hold true.

We try to (or should ?) recognise when we do this that extrapolation is more likely to be in error than interpolation.

15 hours ago, MigL said:

Maybe my understanding of a 'model' is not the same as everyone else ?

We have a 'model' for dams, Studiot.
It comprises the interactions of the stresses and strains, with loads, material strengths, and gravity.
We simply 'plug in' different parameters and conditions to arrive at the five different configurations.

Similarly with climate science, Ken.
We have a model of the atmospheric interactions with solar heating, oceanic heat sinks, and greenhouse gas insulation.
We simply 'plug in' the differing parameters/conditions to arrive at a possible outcome.

The 'model' is the theory of the interactions that govern ( where applicable ) the parameters/conditions we wish to investigate.
For example, the theory of gravity is the model which tells us how different masses, with different speeds will orbit.
While EM theory is the model into which we 'plug in' the charge of a particle and we know which direction it will curve in a particular field.

Or did you mean the scale model of an AVRo Vulcan, which you assemble with plastic cement, paint, and hang in your bedroom.
( yes, I had one of those 🙂 .

 

A model may be physical (hydraulic ones often are) or theoretical as you describe.

Is that a problem ?

[MigL]

No problem at all, Studiot.
Just making sure we're all on the same page.

[geordief]

23 hours ago, studiot said:

No one with any sense will make no attempt to test a model.

It may not be possible to match the model against the actual thing being modelled (perhaps like my dam because it does not yet exist) or perhaps because it pertains to conditions we cannot achieve on Earth such as processes with a star.

But, it makes sense to 'test' the model under simpler known conditions and hope that the extrapolation will hold true.

We try to (or should ?) recognise when we do this that extrapolation is more likely to be in error than interpolation.

 

 

I think my mistake may have been to subconsciously apply the arguments that may exist  around  fundamental models of the universe (GR,QFT etc) to the manifold kinds of models that  others here have shown to exist.

 

Luckily I posted in Philosophy where such mistakes are perhaps  more pardonable.