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

The possible vs the actual

  1. Oct 24, 2009 #1


    User Avatar
    Gold Member

    Many of the debates on this forum boil down to the relation between the possible and the actual. Are they the same size or is the realm of the possible larger than the realm of the actual?

    We see this in arguments over whether maths (taken as a map of all possible worlds) is larger than all actual worlds. So does math = reality or is math > reality?

    The same fundamental question arises with the many world interpretations of quantum theory, with multiverse approaches to cosmology, and with information theoretic approaches generally (as in Turing and Tegmark).

    1) So is anything that is possible, also going to be actual? Why?

    2) Or is there always going to be more that is possible than can be actual? Why.

    3) The third choice of course is more things are actual than are possible. That at least seems a safe one to reject.
  2. jcsd
  3. Oct 24, 2009 #2
    You need to be more clear:


    Actual things := things that exist. Examples are tables, cars, plants etc.

    Possible things := Things that are logically possible: Examples are "golden mountain", tables, chairs, plants etc.

    Note that actual things actually form a subset of all possible things.

    Classical debate( you can read the link i provided below):

    Non-actualist: The claim is that there are non-actual, possible things.

    actualist : The claim is that there are only actual things. There are no possible things.

    David Lewis on the classical debate:

    Any many others( robert nozick, tagmark, lewis) say the following: All possible things are actual.

    read: http://plato.stanford.edu/entries/actualism/
    Last edited: Oct 24, 2009
  4. Oct 24, 2009 #3
    To your question:

    1) you are asking if lewis ` s thesis is right( see post 2). Lewis gave the reason that it is more fruthful to accept his thesis becuase of the reduction of modal notions( necessities, and possibilities).

    2) This is just the non-actualist view( see post 2). It is completely not relevent to lewis ` s thesis.

    3) sure.
    Last edited: Oct 24, 2009
  5. Oct 24, 2009 #4


    User Avatar
    Gold Member

    So possible > actual according to this view.

    And possibility then seems to be aligned with the idea of generality. The familiar dichotomy of particulars and universals. The universal harbours more possibility.

    Agreed this feels the right way to go. But then we get into the difficulty of what is the set of all possible *general things*? Or Plato's realm of form in which there is one perfect example of every possible thing.

    So is a particular table a subset of the general notion of tableness? And is tableness really a "general entity" or instead a "general organisation"? Is it in fact really a particular example of the even more abstract category of "furniture". And if we can keep on generalising our generals to move ever further up this chain, where will we arrive? Some single all encompassing form? If not, why not?

    So your attempts here to define the question in terms of a process of generalisation, and being about thingness, are already running into serious problems.

    You need now to define what you mean by things and by the process of generalisation.

    Thanks for the link but was there some particular position you want to draw attention to here? Which variety of actualism for a start?

    Anyway, the general actualist position would be possible = actual. And it founded on the (mis)assumption that the actual is all about substance. Only that which is concrete and material can claim existence. So aliens in other possible worlds are non-concrete particulars and so not actually possible. The immaterial idea does not qualify as being part of what gets counted.

    Again this raises the status of forms and generalisations. Aliens are a generalisation (a generalised notion of lifeforms). So why would we not keep generalising further? How can we resolve this debate using aliens in other worlds as our substantive argument?

    In fact, even "possible worlds" is smuggling in a substance based ontology. It appeals to material existence rather than formal.

    The actualist camp is attempting to define possibilia in terms of particular objects (or lightly generalised objects). But have we yet ruled out form - organisation, relationships, pattern, purpose, meanings - as what is at least also potential, also a content of possibilia.

    This connects with the INUS approach and others that stress the role of context in causality. Could Winston Churchill have been a german? Well, his parents could have moved there and given birth. But would he really have been "Winston Churchill"? Clearly, much of what we think of as the human that exists is really the locus of a particular historical context, a context with particular constraints.

    So we do then have to be able to generalise the notion of context to answer questions about what is part of the possible. Is the space of possible contexts > or = to the space of actual contexts?

    Difficulties proliferate. And the actualist vs non-actualist reference frame of discussion does not properly define the process of generalisation (how it is done, where it leads), nor does it pay enough attention to the issue of contexts (global constraints) or the putative actual existence of form.
  6. Oct 24, 2009 #5


    User Avatar
    Gold Member

    If your choice is to defend Lewis, then you will have to supply arguments about how contexts and forms can also be reduced.

    And are you thinking that reduction = generalisation? Do you view the terms synonymously?
  7. Oct 25, 2009 #6

    what is > mean?

    I will not comment about universals, because this outside the topic.

    No comment( outside topic).

    Nothing i said said here are controversial. They are a list of definitions you can find at the stanford link. I have no idea why you start talking about universal, particular, and "general things". You need to focus, or i cannot talk to you.

    No. They are just a list of definitions & classical positions from the stanford link.

    How is aliens in other possible world not concrete if those are concrete possible worlds?

    Why is aliens a generalization? If tomorrow, people found life on mars, then we would say these life-form are aliens. Why would we called them generalization?

    What is your definition of possible worlds? Are they concrete, or abstract?

    Give me an example for "meaning".

    What is this "INUS approach"?

    I have no idea what you mean by " space of possible contexts".

    I will say that what is actual is a subset of what is possible, where possible and actual is defined in post 2.
    What do you mean by "process of generalization", and "putative actual existence of form"?
  8. Oct 25, 2009 #7
    Lewis say that by accepting concrete possible worlds in to are ontology, we can explain "possiblities", and "necessities" in terms of quantifications over possible worlds.

    Do you know anything about modal logic?

    Reduction, or reductive is just a fancy way of saying " explain away"
  9. Oct 25, 2009 #8


    User Avatar
    Gold Member

    Vectorcube, clearly you have no interest in serious discussion so perhaps just leave it to others.
  10. Oct 25, 2009 #9
    In fact i am serious all along. I keep technical language to a minimum just so that i can have a decent discussion with a non-specialist.

    As you can see. In all my replies to you, I have only asked for you to be clear, focus, and give examples. I am not used to unfocus discussions. Think of how you would write a book in analytic philosophy, or math. That is the amount of rigor i expect from you, or else, you cannot benefit me in anyway.

    If i wrong you in anyway, then i apologize.
  11. Oct 25, 2009 #10


    User Avatar
    Gold Member

    I forgot to mention the superstring landscape as another possible vs actual issue in physics. So this is a very live problem.

    People are polarised as to whether to expect the space of the possible to be immensely larger than the space of the actual, or whether to think all that is possible will actually, somewhere, have to happen.

    In QM, this leads to rival interpretations like decoherence vs many worlds. In cosmology. it becomes multiverses vs the big bang as a single creation event.

    Modal logic was a recent rather narrow discourse on the dichotomy of possibility~necessity and instructive mainly in its failures.

    So, again, does the space of the possible = that of the actual, or is it >? Arguments rather than links please.

    (Note the mysterious symbol > stands for "greater than" - as any handy eight-year-old math scholar will be able to tell you).
  12. Oct 25, 2009 #11


    User Avatar

    I agree with this characteriztion of most discussions so far. I'll back later with my personal views on this, I have a particular view to this, that introduces a notion of "actual possibility" and this is the tools that should have shave off non-physical possibilities from physical models.

    I'll try to explain later.

    Like you mention already with string theory, there seems to be something pathological about the situation when we basically drown in the set of possibilities. This also connects to computability, inferrability and representation capacity.

    The thing that isn't distinguished between in physics, in particular string theory to take an example is that what's possible is RELATIVE. What possible to imagine for a human is quite different that what's physically inferrable between say two subatomic systems. Once this is acknowledged, I think alot of the mathematical redundancy we have can be reduced, and this makes our navigation in hypothesis space far more rational.

  13. Oct 25, 2009 #12
    There are many errors here that i would not want to comment. Just want to say that the string landscape, QM worlds are not at all non-actual possible worlds. If they are true, then they are part of the actual world.

    Well, that is lewis ` s thesis, and i gave you one of his reason. Namely, you can have a reductive explanation for primitive modality.

    If don` t know, then you merely need to ask.

    Are you sure it does not mean material implication? :surprised
    Last edited: Oct 25, 2009
  14. Oct 25, 2009 #13


    User Avatar

    Mmm... I wonder if we are misunderstanding each other here. We are in the philosophy forum, but the philosophy forum in the physicsfourms - so at least I am under the impression that we aren't discussing some pure philosophy or history of philosohpy in general, I thought we were discussing how this can be constructive specifically in the philosophy of science, and physics in particular.

    Ie. the fundamental problem is that of learning about our world, a strategy for finding the best predictive engine for inferring the future based on the present, etc.

    There are open questions here, this is why to a certain extent the discussion unavoidably are somewhat fuzzy. Different research programs can often be characterized by somewhat different positions on the questions we are discussing here. The string landscape is a good, relevant example IMHO.

  15. Oct 25, 2009 #14

    Well, i made comments for that. The multiverse if true are not possible, non-actual things, but an actual thing.
  16. Oct 25, 2009 #15


    User Avatar
    Gold Member

    Statements are not arguments.
  17. Oct 25, 2009 #16

    1) If multiverse is real, then the multiverse is actual( by definition).

    2) If multiverse is real, then it is logically possible, thus Possible( definition).

    3) By 1 & 2, if the multiverse is real, then it is possible, and actual.

    4) But anything is both actual and possible are part of the actual world.

    5) 3 and 4 suggest that the multiverse is part of the actual world.
  18. Oct 25, 2009 #17


    User Avatar
    Gold Member

    Now what happens when you run string landscapes in this argument?
  19. Oct 25, 2009 #18

    The string landscapes is a multiverse if it is real. It would also part of the actual world. Many people that do not know **** about the matter often confuse multiverse in physics with the possible worlds in modal realism. They think it is the same thing. It is interesting, because if the string landscape is true, then it would only be part of the actual world. Modal realism is a much more general thesis than the multiverse think up by physicists.
  20. Oct 25, 2009 #19


    User Avatar
    Gold Member

    Apeiron – I don’t think you’ve gotten to the basic ontological issue here. The problem here isn’t form vs substance – it’s whether possibility should be conceived from the beginning in terms of what’s actual.

    “Actual” essentially means, what’s given – i.e. what we can take for granted in the world. And from this we can develop various notions of possibility, i.e. what’s not actually given, in reality, but might in some sense be given... logically, or in the future, or in a realm of pure mathematical form, or an alternative cosmos, etc.

    There’s nothing wrong with that, but I don’t think we’ll find much that’s interesting to explore down this well-trodden path... because the key issue is not being considered. It’s already been implicitly resolved in favor of what’s given... whether "givenness" is conceived as form or material substance, etc.

    We only get to the real issue of possibility when we ask how things are “given” – where does “what’s actual” come from? As you know, the traditional idea – “causality” in its many forms – is that what’s given comes from something that was given previously (“initial conditions”), and on what’s ultimately given (e.g. “laws of physics” or “principles of logic”). The given actuality is taken as fundamental, and possibility as merely a derivative notion.

    Now it’s easy to assume that something must just be “given” at the bottom of things – i.e. there must be something we just have to take for granted. Back to the earliest philosophers, the goal of philosophy and science has always been to identify this ultimately given reality. But going almost as far back, in our tradition, is the sense that there’s something else going on, in the dynamics of how things come into being, that shouldn’t be taken for granted.

    I know you are very alive to this undercurrent in the intellectual tradition – at least, that’s how I understand your going back to Anaximander’s apeiron and your description of what’s ultimately given as “vagueness” or “pure potentiality”, in other threads. To me, this kind of language points to a deep challenge that philosophy has yet to meet. Essentially, we want to describe a world in which how things become possible is at least as fundamental as what they actually are, as given fact.

    Now we already have, in evolutionary biology, a well-established science that works in these terms. Practically speaking, in biology “causation” has been entirely replaced by the notion of the conditions under which random accident can give rise to complex structure. And there are many indications that a similar approach is needed in physics and cosmology. But IMHO, philosophy has lagged behind. Heidegger made a strong beginning in Being and Time, but since then there’s been little or no progress toward an ontology in which the structure of possibility plays a fundamental role.

    You’ve been eloquent in other threads on the “anticipatory” nature of conscious awareness – which was also a central theme in Being and Time. We can be conscious of “what’s actually there” in the world around us, because our brains are constantly projecting “what might be there” in advance. So in a sense, our deepest layer of experience is “made of” possibility – as opposed to the traditional data-processing model of awareness.

    I think of existence as a dialectic between possibility and actuality, each with their own distinct structures. Things can only become actual in a context of relationships that makes that actuality possible; and then what they actually become makes new relationship-contexts possible. “Possibility” in this sense is very different from “logical” possibility – it’s a fundamental aspect of the complex concreteness of what happens in this world as it evolves.
  21. Oct 25, 2009 #20
    What has this to do with logical possibilities?

    Not ture. Something might be a priori, but false, because it is a posteriori necessary true.
    If you don` t understand, then you merely need to ask.
  22. Oct 25, 2009 #21


    User Avatar
    Gold Member

    I agree with pretty much everything you say. I would indeed redefine the possible (usually taken to mean logically crisp variety) in terms of a vagueness, a pure potential.

    But substance and form are useful to focus on here as they are both generalisations - in complementary directions - which emerge together out of vagueness. As you say, we want to look at what comes out, what is real around us, then infer something about the initial conditions of reality. And if we find a world sharply divided by the dichotomy, the asymmetry of substance and form, then working backwards to the vague beginnings, substance and form will eventually blur back into the same foggy state of nothingness/everythingness that is just a simple potential.

    So this is the modern reductionist story vs the vagueness-based alternative.

    1) crisp local variety (a collection of atomistic microstates) => emergent macrostates

    2) vagueness (a symmetric mix of micro and macro potential) => the dichotomised reality of local substance~global form.

    Modal logic would be an example of a discourse which adopts the first formula and so cannot "see" the second.

    So how does 2) fit with the question I originally posed?

    Is the space of the possible larger than, or equal to, the space of the actual?

    You can see that with 1), we are trying to answer things by counting microstates (even if in many worlds and multiverses and stringscapes, we seem to be treating whole worlds as microstates). Which is where things really start to get illogical. In what way can worlds be countable microstates? The premise basic to the modal logic argument is philosophically flawed.

    But with 2), the space of the possible is both unbounded (limiteless) and also smaller in some useful sense. The vague is just one kind of thing (so less than the two kinds of things - like substance and form which can separate off from it). Yet it is still larger in the sense that all is still possible (which is no longer true once some dichotomised state becomes crisply actual).

    In biology, a seed could grow into a tree with an infinite variety of branching patterns. But once a tree has grown, it is stuck with its history of branching. So were those microstates actually "in" the seed as countable variety. Or were they just a smear of vague potential?

    These are frontier questions in philosophy - for philosophers who are responding to the advances in physics, biology and branches of math like chaos theory and dissipative systems.

    Anyway, you are very correct in saying this is an anthropic and observer-including approach to possibility and initial conditions thinking.

    We have to look at what exists/persists around us - deal with what seems actual - and then project back, generalise, to gain some model of reality's causal origins. This is actually a constraint on our philosophising.

    We should imagine no more possibility than what is required to generate our perceptible reality!

    So we want to make the actual = the possible. We want a conservation of what exists. A first law of causality. Yet we also need to leave room for the fact of development, a second law of causality in which there is an arrow of time, a direction things go.

    So we also want fundamentally to be able to say that possible > actual. That there is an "entropic" slope down which things can run so that change is also natural to the world.

    This is the kind of subtlety of thought that is quite beyond academic philosophers still stuck in Newtonian modes of modelling. They have yet to catch up with the second law of thermodynamics. But it has been an issue for decades among my friends who are theoretical biologists for example.

    Anyway, we now have a more complex view emerging.

    We want the actual to equal the possible in the sense that whatever emerges must be quantitatively a match for what was originally there. If our theory of existence exhibits a principle of conservation, then we know that it is "good". There is a deep reason for requiring our notions of causal origination to show a conservation in some fashion.

    But we also must require that our fundamental model has the possibility for change, for development, for qualitative difference. So there must also be some version of a second law, and entropic slope, wired into the causal model, the logic that we think underpins reality as a whole.

    The space of the possible is then neither really large nor smaller than the space of the actual. It is larger in having more that is potential. And smaller in being undiffentiated and so not yet separated in at least two directions. Which is where we can see that neither > nor < are symbols that really serve our purpose here. Instead, we may begin to see more deeply what a vagueness actually is. A realm where both larger and smaller have lost their difference, lost their distinction. A realm which is both, and thus neither.

    Logical possibility is about the atomisation of form. Seeing reality reduced to information or microstates. This is a useful way to model things because it is simpler. But it relies on us humans filling in the background with our own understandings about observers, contexts, histories, anticipations, etc. The equally crucial stuff that gets left out of the formal model based on crisp local variety - atoms, information, events, possibilities, trajectories.

    So ultimately, what gets left out of the conventional modelling becomes mysterious to people.

    We have to go back and fix things by including both the local and the global, the substance and the form, the stasis and the change, in the modelling - even the modelling of logic or causality. Or rather, especially in our logic/causal models.
  23. Oct 26, 2009 #22


    User Avatar

    To the original topic here is my personal take on it.

    After some thought I decided to skip writing a lenghty lineout of my own general reasoning and instead do a simplified comment and just jump right at the point and project my own thinking onto the terms you are discussing, although it might not be my own choice of terms.

    This is how I see it:

    - To me, both actuals and possibilities are observer dependent.
    - The distinguishable actuals for one observer is the things this observer thinks are facts, or limiting cases of distinguishable possibilities where a highly probable possibility is simly indisitinguishable from a fact (note that this doesn't mean the it's universally or eternally true)
    - The observers action follows from a rational action conjecture, from the observers facts and possibilities. So the facts and possibilities then get predictable consequences, it's not only metaphysical notions, the physical action is constructed from this fact/possibilitie system (I call this system the observers system of microstructures).

    So in my view, in a certain sense all possibilities are actual, BUT the key points are:

    Both actuals and possibilities are constrained by the observers complexity, this is why there is no such thing as a mathematical infinity of possibilities. Or rather THIS "mathematical possibility" is not something that influences the ACTION. Only the distinguishable actual possibilities influence a systems action. This is the way I envision that this idea is going to make a difference to real modelling.

    There is always and observer, and it's environment. The observers own structure are the actuals, and the unknown environment are only "possibilities" in the observers microstructure. This is what I mean with "actual possibilities". The difference between a possibility and an actuality, is like the difference (in my view) between a possibility and a fact. And when the uncertainty of a possibility is indistinguishable, it's indistinguishable from a fact, form the inside view. But even a fact can later need to be revised, because there are no timeless facts for the reason that there is no physical infernece system, that can make a certain timeless 100% inference (deduction).

    This is why "logical possibilities" to me doesn't refer to deductive logic, the refer to a more geneeral inference, the closes choice is inductive inference/logic.

    So the notion of ACTION is what connects actuality and possibility in my view.

  24. Oct 26, 2009 #23
    That is confusing for me, because "possible" means "logically possible". I have no what "vagueness, a pure potential" mean.

    How about some examples?
  25. Oct 26, 2009 #24


    User Avatar

    To take a human example, the actuals are the facts that this human has inferred from interaction with it's own environment throughout it's history. Ie. it is from experience we have come to the conclusion that there are regularities and not yet questionable facts about nature. For example one could think that knowledge of the laws of nature are such actuals. But konwledge of this law is implicitly encoded in the state of this human.

    Thus the actuals for a human is what this human has inferred to be pretty much correct. It's somehow part of this humans rational beleifs.

    It's not hard to imagine that two humans that have totally different experience, that has grown up/evolved in different environments, would come to completely different sets of actuals.

    So the actuals are dependent.

    However, these actuals insufficient to deduce the future perfectly, they only provide a basis for guessing the future, and acting as a guide. So the possibilities are induced from the encoded actuals, inductively determines the observers actions.

    Also another prediction is that when two observers that are in disagreement upon actuals are put in contact, an interaction force appears that forces them to negotiate and equilibrate. This conceptually explains why the world is coherent and everyone (effectively) agrees upon actuals at some level (physical law), although at the more basal level there is not consistency hardcoded. It's an emergent consistency simply because the non-consistent version is not stable, it would self-decompose into consistency, and it's just a matter of timescales.

    To take anothre simplified human example, consider a poker game. The rational player conjecture assumes that each player makes his moves (action) as per his own optimal inference. So the action of the othre players depend on what they know, or what they think they know about what the othre players know. Part of the game becomes not only to choose your own cards, rather an important part of the game is to try to guess what you think the actions of all other players are. This gets predictable consqeuences for such a game. Similar idea are used to mode economical systems.

    Of course, it's not possible to tell exactly what the optimal inference to the inside observer is, but it's enough to CONSTRAIN the probable options, and to place your own bets more wisely.

    A physical example would be two subatomic systems interaction. When you evaluation the feymann path integral for the action, the choice of the structure of paths to be summed are quite important. By simply postulating the structure of possibilities at will, you can pretty much get any result you like. It's also the common problem in physics that these integrals leads to absurdities, like infinites. Apparently this is a sign of the fact that the space of possibilities used in the evaluation of the action contains non-physical redundancies, and this is the core of the problem.

    This is work in progress so don't expect me to solve all the open problems of physics in this thread but the idea is that once you acknowledge tht the action is relative, this constrains the set of PHYSICAL paths in the feymann integral. My idea is that these constrains come from the complexity of the observing system (realising the action), and this is expected to loosely related to it's mass. This means tha the proper "normalisation" must be implemented properly from start. Once you let the redundancies in, you loose track of how to remove them since there is a large ambigouity in renormalisation. There are different mathematical ways to renormalize away infinites but they all give different reuslts. The question is of course, which results that correspond to reality.

    So the phsyical example still says that the actuals are system dependent, because it is subject to inferrability or observability, and the context - the inference system / the observer - obviously matters here. Moreover a given set of information, yields different distinguishable set of possibilities, also constrained by the observers complexity.

    In particle physics the action, is usually formulated relative to a very specific context. It's the laboratory frame, where the particle accelerator etc are all included. Basically the entire environment of the collision domain, including preparation phases.

    What I'm suggesting that we should look for, is the inside view of these things. This will I think lead to new insights of the makeup and hierarchies of interactions and particles.

    I don't have all the answers but this suggest a direction for further research. A direction that is previously neglected. There are some various new ideas coming up. Smolin is tangenting this, this is why I think some things he says should be taken seriously. But there are also others, working on inside views (Olaf Dreyer) and also physics from inference (Ariel Caticha).

    I am working on a combination of physics from inference + evolving law + inside views

  26. Oct 26, 2009 #25


    User Avatar

    I can buy this phrasing.

    But then what I mean with them beeing observer-dependent or relative. Ie. the possibilities are RELATIVE to the choice of logical inference system. Moreover do I suggest that the logical system iteself evolves.

    I think you think of logic as some external universally valid mathematics, that can always be applied. This is where we differ I think. I think this eternal reference is the root of several problems.

    The alternative is evolving logic systems. and nature then consists of several interacting logical systems, and the question of physical law then becomes that of the population of inference systems in nature. The common DNA of these inference systems then correspond to an effective laws. The variation of the inference systems implies a specific set of interactions.

Share this great discussion with others via Reddit, Google+, Twitter, or Facebook