Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Constructivist Philosophy and Classical Thermodynamics

  1. Apr 11, 2010 #1
    Can anyone tell me how a constructivist philosopher would view concepts in classical thermodynamics, say like the state variables pressure, temperature, internal energy, and entopy?

    I read in Wikipedia that "Constructivists maintain that scientific knowledge is constructed by scientists and not discovered from the world ". These days most physicists regard such variables as quantifying macroscopic emergent phenomena. I suppose that from a constructivist perspective thermodynamic phenomena and their quantitative descriptions are invented rather than discovered. But maybe this is wrong?

    I'd like to know what the views of folk in this forum are.

    I've chosen thermodynamics because it's a closed subject in physics that many are familiar with, which was developed long, long ago. Arguments about it's nature are, I hope, not likely to raise the hackles of active researchers in thermodynamics, of which I suppose there are, by now, very few, if any.
  2. jcsd
  3. Apr 11, 2010 #2


    User Avatar
    Gold Member

    I would start by making the commonsense point that either extreme constuctivism or naive realism is going to be wrong. We model the world. So there is some actual formal relationship which involves a constructed model and its "real" measurements.

    Is thermodynamics a good example of this at work? Yes, of course.

    What may be interesting about your particular question is that the macrostate - such as temperature - would typically be considered "more constructed", while the microstates of the system would be "more real". The world is always in some particular state (it is not vague), but we cannot measure that, so we just measure some average notional quantity and give it a name.
  4. Apr 11, 2010 #3
    Interesting that you think the constructivists would view the microstates as more real apeiron. I would think that constructivists would view atoms as being "less real" and perceptions and experiences such as temperature as being more valuable. I mean we all feel hot and cold right? If I remember correctly, Mach was one of the last supporters of this view of atoms as theoretical constructs used to describe phenomena. Brownian motion kind of compels us to take this view of atoms banging into each other and giving rise to temperature more seriously. Since the macrostates are what we observe directly and experience, I would think the microstates are just constructs needed to explain the averages we observe.
  5. Apr 11, 2010 #4


    User Avatar
    Gold Member

    I wasn't expressing my personal preferences here! Not at all.

    Rather, I was offering a view on what I believe is the standard modern materialistic perspective - that the microstates (as in a random arrangement of particles) is what is ontically real while the macrostate (as in the temperature) is merely emergent, which in reductionism has the connotation of being less real.

    I of course challenge both ideas (while remaining still a modeller - so always feeling rather baffled by the heated deates between naive realists and pomo-style social constructionists).

    So for example, the idea of ensembles of discretely existing microstates is a model which I would challenge with the developmental notion of vagueness.

    Likewise, I would see the emergent macrostate as just as "real" in the systems perspective - the global scale has downward causality. Of course, temperature is not really a macrostate in this broader sense. It would be a description that includes all the necessary boundary conditions, such as the container that constrains the collection of particles, and the entropy gradient across its walls that results in some particular internal equilibrium.

    But I was just sticking to the mainstream view of matters in my reply. I just wanted to point out that for thermodynamic realists, the microstates would seem to be more real than the macrostates (because all the causality is in the microstates). And even for a thermodynamic constructivist, the microstates would probably seem to be less constructed than the macrostates.

    BTW, there are still thermodynamic researchers who rail against the Boltzmann ensemble models of entropy, arguing for a caloric interpretation. So the tradition of not believing in discrete atoms but continuous materials lives on.
  6. Apr 12, 2010 #5
    I agree fully with this common sense.

    But another reason for my considering thermodynamics is that it is in a sense the firm ground above a slippery slope.

    With thermodynamics 'extreme constructivism' does seem to me more appropriate than 'naive realism'. But once the idea takes hold that familiar physical variables like temperature may be regarded as (very useful) inventions rather than 'real', the door opens to extrapolating this conclusion to less familiar situations, like the microworld of atoms, and arguments about whether these entities are more (or less) invented (or real) than temperature:
    . Before treading further on this slippery slope it seems to me that it would be useful to settle whether useful "models of the world" , like thermodynamics, are simply inventions which may or may not be isomorphic with 'reality' (whatever this is) but that pass the testing cycle of observation, prediction and confirming experimentation.

    Which of course would rules out much of modern theoretical modeling.
  7. Apr 12, 2010 #6


    User Avatar
    Gold Member

    What is the epistemological conclusion you are driving towards here? The above is rather enigmatic.

    The correct position I feel is that models are constructions. And they can be so simple, so generalised, as to be quite unreal (particles as dimensionless points, etc)

    But then measurement selects among an ensemble of models so that a successful model is never "just a construction". It has to prove itself via prediction and confirmation. A model becomes favoured because of its capacity to serve its modeller's purposes.

    Surely "modern theoretical modelling" has this form?

    Being "isomorphic with reality" would not be an acid test of a model (despite a widespread public belief that science is about truth, about realism). Science is based on the philosophy of pragmatism (yes, CS Peirce again), And the acid test of a model is its utility, its ability to achieve its purpose.
  8. Apr 12, 2010 #7
    I'll try to be clearer. First, I agree that you have stated the correct position and that the simple idealisations you mention don't affect it. And that the pragmatic view is that a model should achieve a purpose. Thermodynamics seems to fit the bill here very well, but I can't see how thermodynamics can be anything else but "just a (very successful) construction". Or model, if you prefer. It certainly has proved itself via prediction and confirmation ---no one has yet made a perpetual motion machine.

    But I'm unsure about modern theoretical modelling, much of which has difficulties with prediction and confirmation.

    For instance take an example: the application of thermodynamics to theoretically model gravity as an 'entropic' force. It has attracted much recent attention, and retro-dicts all sorts of hitherto mysterious stuff. But if temperature and entropy are invented human constructions, then what kind of an animal is gravity?

    Or is it simply our descriptions of gravity, either via Newton, GR or thermodynamics, that constructivist philosophers have in mind as inventions, rather than gravity itself? If so, I'll join their ranks and forget about reality. Except when trying to leap over tall buildings.
  9. Apr 12, 2010 #8


    User Avatar
    Gold Member

    Clearly gravity is a mental construction in the modelling sense. Some want to view gravity in terms of a quantum field theory (so the imagery of what is "real" would be little virtual graviton exchanges), while GR sees gravity as the warp in spacetime (quite a different mental image).

    The angst over what is "real" usually boils down the urge to see things with our own eyes. We want a mental image with perception-like veracity to convince us that our constructs (such as particles, waves, warps, whatever) are a true picture of microscale reality. We feel that if we could have a completely convincing image in our minds, then we would feel satisfied that the real world is understood, comprehended, finally accounted for.

    But while mental imagery (such as particles, waves, warps, whatever) is always useful to theoreticians to get theories started (and to teach theory to neophytes), the images must themselves fall by the wayside ontologically, being replaced by the formal models which encode a modelling relationship with the world.

    This is the pragmatic epistemological position that science takes. You can come up with some new cute idea like reality being composed of writhing string loops and see whether such images leads eventually to a concrete system of modelling and measurement that serves a purpose.

    There should be no intellectual angst over the obviously constructed nature of the mental images that motivate physics, or other sciences, as the freedom to be quite unreal (I mean, wriggling little loops of string?) has often proved to be the way to make big advances in theory.

    The classic case was Newton just accepting action at a distance as a modelling assumption in his law of gravitation, when Descartes and others were bogging themselves down with the more "realistic" insistence that a motile sea of atoms must somehow be propagating the swirling motions of celestial gravitational force.

    Of course, field theories and geodynamic imagery later came along to allow yet further advances. But the path to effective modelling is not necessarily the path to a more veridical faux perception of the scales of existence beyond the normal range of our conscious experience.

    And we of course know that even our conscious experience is subjective modelling rather than a direct experiencing of objective reality. So even the most convincing mental images are always still only constructions in the pragmatic modelling sense.

    Who feels angst that they look at the three phosphor system of a TV screen and yet believe they are experiencing "full colour"?

    Really, you are only seeing three very narrow wavelengths of visible light. The screen is not emitting all the frequencies you think you see like the naturally-lit world around it.

    But do you care that you cannot tell this real difference? That your experience is in fact so clearly constructed? That an alien with a differently evolved visual system would stare at your TV and wonder why a scramble of signal could hold your attention so long?

    Well, the same level of angst should apply to scientific models of reality and their constructed nature.

    Well-constructed things work and badly constructed things fall apart. What more is there to say?
    Last edited: Apr 12, 2010
  10. Apr 13, 2010 #9
    I liked this last post of yours very much, Apieron. Good common sense all the way through.
    Thanks for writing such full replies. They’re appreciated.

    Just one explanation, hopefully not too enigmatic:

    I have for some time thought that many mathematicians don't understand what they are doing. I don't mean this in a nasty disparaging sense, but just as a comment to be taken at face value. The mathematicians I have encountered all seem to believe they are discovering a world of eternal Platonic truth: they think that 2 + 2 = 4 was true long before humanity existed. Whereas to me it seems obvious that maths is a very clever, effective, invented construction, a distant cousin to music and chess, rather than eternal truth.

    Because thermodynamics is such an egregious example of an invented story couched in invented terms (like temperature), which on other scales are quite inappropriate, I then began to suspect physicists of a similar fault, namely of believing that they are inventing models that approximate ever more closely to a presently still rather mysterious 'real' world, which somehow (fortunately) just is. An their final goal, an ultimate unified and accurate Theory of Everything, seemed to me unattainable while parts of it were just special inventions like thermodynamics.

    But, as you have explained, model building is an eminently practical activity (it also keeps
    physicists off the streets). It may nevertheless be helpful to recognize what we are: just a loquatious species of African apes who have invented a sophisticated language (maths) to describe in various ways --- predictively and therefore usefully --- the circumstances we find ourselves in.

    I liked your last comment particularly:
  11. Apr 13, 2010 #10
    Any time you measure something, you use a chosen metric, scope, and usually have a purpose.

    Knowledge is not 'data'. And even data is just a description.

    Scientists choose what to study, and they build their knowledge on a theoretical framework which has a history, and a place in history, and society.

    Zero degrees celcius is the freezing point of water.... under predefined conditions.

    Its not so much that scientific knowledge is constructed... but that all knowledge is. Knowledge of a rock is not a rock.
  12. Apr 13, 2010 #11


    User Avatar
    Gold Member

    As is being argued in the adjacent Hardy thread, there is actually something "more real" about certain kinds of mathematical truth, like pi, e and lie algebras, and also physical "truths" like the speed of light, zero degrees kelvin, and other constants of nature.

    So could we say that pi would be pi in any possible world (of flat dimensionality of course)?

    And maybe even absolute zero would be the same lowest temperature for any possible world? (Though of course, a world with a smaller planck scale could get that much nearer to absolute zero).

    There would be certain patterns of nature that would seem to be invariant, no matter how a world may be constructed. All that is required of the world is some global self-consistent uniformity and certain local facts would then always emerge the same.
    Last edited: Apr 14, 2010
  13. Apr 14, 2010 #12


    User Avatar
    Gold Member

    [Furthermore]: The interesting possibility here is that precisely two things can be certain - existence and symmetry.

    The Cartesean cogito ergo sum says the fact we cannot doubt, the fact that assuredly is not a construction, is the fact we exist, and so something exists.

    The principle of symmetry (of invariance, of equilbrium) may also perhaps be a second certain fact (and it would be the fact at the heart of maths and other descriptions of the fundamental patterns of reality).

    Existence (which we have accepted as fact) is what survives change. And symmetry also is what you are left with when all possible changes fail to create an actual change. Turn a circle as much as you like and it still looks the same.

    Change is what makes statements about existence doubtful (things could be otherwise). Symmetry then describes what must survive any possible change, and so becomes beyond doubt.

    This would be a Platonic line of thought perhaps. But the symmetry principle is clearly fundmental to both physics modelling and the "purer" mathematical modelling of self-organising, or self-consistent, pattern generally.

    So we cannot doubt that we personally exist. And perhaps we cannot doubt that symmetry exists as the most general principle. For there to be existence, there must be something that can survive all possible change. Which in turns says symmetry is truth because, by definition, it is what (we find) survives all change.

    This is why people can feel 1+1=2 in all possible worlds. (The integers being derived from symmetry principles - a further argument lies here as to how....). Certainly pi, e and other mathematical constants are symmetries (eternal ratios) that would seem true of all possible worlds.
  14. Apr 15, 2010 #13

    Good thinking indeed!

    My contention is that there is more in this universe than matter and energy. Consciousness is one of them.

    Can consciousness be explained in terms of atomic or sub-atomic processes?

    Hopefully, I am not too vague.
  15. Apr 16, 2010 #14
    This reveals a slightly less than commonsense touch, Apieron! Symmetry is an invented operational concept.

    To demonstrate that something can be described as symmetric you do something to it ---- perform an operation on it --- and then compare the result with the original; another operation. Both doing and comparing are an invented procedures. As in the example you (or maybe someone else) invented: "Turn a circle as much as you like and it still looks the same".

    There is also the complication that a procedure involves the passage of time as well as human intervention, and sadly we have no idea (or perhaps too many ideas) of what "time" is all about.


    Or perhaps we can!
  16. Apr 16, 2010 #15
    Thanks Deepak. I think this is a very deep topic which philosophers can only argue about. But don't take what they have said too seriously --- we have only recently been made fully aware (say over the last 50 years or so, by palaeontologists) that we are 'just' an extended tribe of African apes. We chatter, speculate, argue and pontificate because evolution has (very advantageously) made us compulsive story tellers. But much of non-evidence-based wisdom that has been placed on the record by past thinkers must needs updating, don't you think?
  17. Apr 16, 2010 #16


    User Avatar
    Gold Member

    It is only an undeveloped thought so I won't try to defend it too strongly here.

    But there is something most people find convincing about the undoubtability of certain mathematical truths. So it is a little bit commonsensical to many that some general facts of existence must be because they cannot be doubted.

    I think symmetry then cuts to the essence of this. So my argument would run, we cannot doubt our own existence of thinkers. Then using that thinking to imagine all kinds of possible difference or change, we will eventually be left with the residue of what does not change, what makes no difference.

    So we have the thinker and also the most enduring possible thought. Symmetry is not something invented but instead found as the result of conceptual doubting - imagining all possible changes and transformations, all the way things might be different, and then observing what is left as the result of this mental operation.

    As to the operation of thought taking time, it is true. But how does that not apply to the procedure of doubting that leaves in the end only the certainty of the existence of the self?

    Are we not talking here of two directions of doubting - so the inner-directed one that leaves only the self certain, and the outer-directed one that leaves only the deep concept of symmetry?

    But again, it is only a stray thought that I found interesting.
  18. Apr 16, 2010 #17
    This is a pretty sweeping statement, but it makes good sense. And constructed knowledge depends on what you are. To an unknown extent knowledge must be species-specific. To paraphrase Margaret Atwood in her gem of clarity, Payback: knowledge about say, thermodynamics is located on a smorgasbord of topics with a sign on it reading Homo sapiens sapiens. This knowledge isn't on the smorgasbord labelled Spiders where knowledge about spinning webs to catch flies is. Tastes differ and so do constructed flavours of knowledge.......
  19. Apr 16, 2010 #18
    I include myself with 'most people' and agree that common sense suggests that we cannot doubt our own existence as thinkers...... but what exactly do we aim our outer-directed doubt/acceptance of existence at? Why pick on symmetry? Is it because it's a notion that is easily contained in one's own mind, and therefore simplifies description for us, like the simple arithmetic of 1 + 1 = 2?

    Or perhaps picking is just a matter of choice. The more I live, read and communicate, the more I choose to accept that there is indeed a physical world existing out there; one in which evolution has generated in us animals a driven need to tell stories. Perhaps one should try to separate in one's mind the stories from the physical world itself, as Joe Dawg suggested above. The content of stories probably depends on what we are (people or say, spiders) as well as on the still-mysterious nature of the physical world. And those that are evidence-based and successfully predictive endure, while others fall apart, as you've said.

    Symmetry helps us enormously in constructing models of physical things --- modern theoretical physics could hardly exist without this notion --- symmetry is the usual first crutch we seek in modelling complicated situations, as in the GR description of matter super-compacted by gravity. It is also a satisfying attribute of beauty and provides balance for symmetric structures in a uniform gravitational field, but this is a matter for other cultures. One thing about which there is no doubt is that that symmetry is a hugely important feature of the stories we tell --- of the constructed knowledge we devise. Could it be just a feature of the physical world that we find so striking because it assists us to the lazy ease of simplification?

    To take us back to a mundane example of constructed knowledge --- thermodynamics. It seems to me that this is a beautiful and useful story of how emergent concepts are related in a mathematically symmetric way as pairs of conjugate state variables like temperature and entropy. But do such invented emergent concepts 'exist' as you argue symmetry does? Or are they all just convenient inventions?

    "Stray thoughts that (you find) interesting" certainly seem so to me.
  20. Apr 16, 2010 #19
    Yes, and this updating should be in conformity with the facts of modern age as well as should try to shun the repetitive parts.
  21. Apr 16, 2010 #20


    User Avatar
    Gold Member

    My use of symmetry is perhaps rather broad here. I don't mean just rotations, transformations and the usual introductory examples. I mean also fractals (scale symmetry), equilibria and other symmetric phenomena.

    My definition of symmetry is essentially a dynamic, emergent, one. Symmetry = a state where changes fails to make a change.

    Rotating a circle is a very simple example. A thermodynamic equilbrium is a more complex one. When all the particles of an ideal gas dash about, the temperatures and pressure still remain the same because there is a basic symmetry in the distribution of the microstates.

    So symmetries would be invariances. And philosophers like Robert Nozick have written about how subjective invariance is our baseline for what we can take to be objectively true.

    Then having established a symmetry, we can then derive further "truths" as symmetry breakings. From imagining extreme asymmetries, such as the figure~ground, event~context, style breaking of the one and the many.

    It is the primitive notion of the object in a setting that led to the numeral 1 and the idea of counting. It was not god who created integers but a symmetry-breaking of reality by perception which proceded all the way to its most extreme possible state (its minima) of the one and the many.

    Then of course came the further symmetry breaking which produced the notion of the everything (that exists) and the void. The 1 and the zero. Or later, zero and infinity.

    So all the certainties of maths (and metaphysics of course) derive from notions of symmetry (what cannot be doubted because any imaginable change leaves them unchanged) and then the further truths derived from the breakings of these symmetries.

    [edit]: though of course we usually discover them in reverse fashion, noting we exist in a broken world with objects and contexts, then working our way back to a conception of the unbroken state from which such asymmetries of nature might have developed. Exactly the story for the standard model and fundamental physics.

    I, of course, argue that symmetries = vagueness, and symmetry breakings that are maximally asymmetric = dichotomies. And what is divided can then mix to establish a new equilibrium symmetry state, which = hierarchies.

    Or in the terminology of Peircean semiosis (his logic of vagueness), firstness, secondness and thirdness.
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook