Critical Realist Epistemology
Offline Discussion moved from: http://www.rationalresponders.com/forum/13114
First off, thanks for the very lucid and thoughtful response. I hope I can make some sense here too.
I'll start with a disclaimer that may be obvious, but it's worth saying anyway: this is just
my understanding
of Critical Realist epistemology, and Polanyi, and Wright, etc. I could be totally misconstruing what they are really saying. Probably am 
Quote:"Critical Realist" epistemology is actually the preferred epistemology of most working scientists, from what I understand. Bertrand Russell is even included among their ranks, if I'm not mistaken. I think it can basically be summed up by saying that while absolute certainty is impossible, the pursuit of knowledge is not necessarily a fruitless enterprise. I think Michael Polanyi is an important figure in this line of epistemology, who points out that presuppositions are the necessary tools we use to work out a picture of reality; the only way to examine anything is with the "tool" of a presupposition. If you want to think critically about your presupposition, you can only do it by using yet another presupposition. As such, Wright suggests that "The proof of the pudding is in the eating."Ok. Not what I was remembering. It sounds as if this is going to end up at the supposed "problem of induction," which is addressed really well by probability (Bayes). I say addressed because there's some quibbling over whether or not the problem is solved or circumvented, but for my money, if you can legitimately circumvent a problem, you have solved it....
I'm really not familiar with Bayes, or the "problem of induction"; I'll have to read up on that.
I don't know too many scientists who would even use the phrase "the pursuit of knowledge is not necessarily a fruitless enterprise." I think it would be more like "The pursuit of knowledge is the only way to effectively gain specific knowledge." The descent into nihilism negates the argument that uncertainty = no knowledge. The only question is whether or not we can come up with an intermediate definition of knowledge between deductive certainty and complete ignorance. The traditional "justified true belief," where justification comes from empirical scientific certainty, and true comes from accurate data, seems to do just fine.
If these folks are not going to end up at the problem of induction, how are they going to get to even empirical plausibility for god? If they're not going to get to empirical plausibility, how are they going to justify knowledge?
Right on. Critical realism is the attempt to arrive at just such an intermediate position, but Polanyi, I think, expands the concept of justification beyond only "empirical scientific certainty", with what he calls "tacit knowledge". I think the reasoning goes something like this:
If presuppositions are unavoidable (as I mentioned before), and cannot be justified without presupposing something else, then it must be the case that if we have "knowledge" (i.e. justified, true belief) at all, we must "know more than we can tell." As such, when dealing with ultimate, foundational presuppositions, justification (in the traditional sense) is not necessary to begin with. Justification comes not before assent to the presupposition, but after you sort of "test drive" the belief to see if it works or not. "The proof of the pudding is in the eating", as they say.
As for presuppositions, I think this is adequately addressed by noting the difference between individual colloquial presupposition and the existence of justification. I do not need to presuppose my own existence, for I experience it directly.
I think I would agree with you that you can have certainty about your own existence, and mathematics. But someone pointed me recently to the fact that an Eastern philosopher might have a problem with Descartes' cogito, and foundationalist certainty of your own existence, because according to some Eastern thought, the ultimate truth is "No self". I don't really understand this, but I'm interested in trying to understand it better.
From there, if I take the time, I can deduce logic and math entirely for myself. With these tools, I can deduce probability theory. At this point, all I have to do is avoid nihilism by admitting that what I perceive is a mental representation of an actual reality, and I'm home free.
I think that's where the problem comes in: such an admittance is an unjustified belief. What is more: we know from experience that what we perceive can be deceiving.
Plug in what, by all available standards, appears to be empirically true, and trust probability, which is deductively derived.
There's a problem with probability too, because reality sometimes (obviously) contradicts probability. As such, you can have justified knowledge (via probability) that is not true, such as my justified belief that the air conditioner would turn on (like it has turned on every other time: probability) the other day when I pressed the button, but it didn't because it was broken. Or to use the BIV illustration, it is improbable that I am a brain in a vat, but it is still possible that I am. If it is the case that I am a BIV, then I have justified, false belief.
Now take another case closer to our ultimate issue. According to science and medicine, it was improbable that the cancer Joy Gresham Lewis' bones would get better when she seemed to be on her deathbed. But C.S. Lewis believed that when he prayed that he could suffer in her stead, she would get better, and he would get worse. His belief was not justified in the empirical-scientific sense, but his belief was "true" because that is what happened.
The theist suggestion would then be (I know you've heard this before... it probably even has a name to it) that while belief in God may not be justified in the scientific sense, it is one of those instances that the true belief is the one that appears improbable (according to traditional terms). This is a presuppositional belief to which you assent in order to find its justification. As I mentioned earlier, so far for me, assent to Christianity has made enough sense out of the reality I experience that I don't see the need for changing my belief.
To use still another illustration: syntax. It is sometimes the case that in a sentence, you cannot know the true meaning of a word until you reach the very end of the sentence, or sometimes until you reach the end of the paragraph, or whatever the case may be. You stick with your initial "hunch" of a definition until that hunch proves inadequate. But even then, you can take a "risk" and keep your original definition. It may end up correct in the end.
But really it is a risk whether you keep your original definition, or change to the more probable one. How many times have you got into what you thought would be the faster checkout lane because there were less people, and less items to be checked, only to find out that this woman had 30 coupons stuffed into her pocket she was going to use?
But of course, then you have the problem of basing belief on what is improbable. But I think you could say that the improbable belief is not arbitrary: there are reasons for such a belief.
Ugh... it is 3:30 am... there is much more that needs to be said, including the obvious fact that I do not have all of this neatly worked out; I'm still trying to understand this Critical Realist mumbo jumbo. There is probably also much that should be edited out of this response...I am sorry for not being as lucid! Must sleep now....
Ockham's Razor is only as sharp as you are.
- Login to post comments
But we can't possibly have any knowledge other than that which we observe or derive from our experiences. Furthermore, even though our initial perceptions may not accurately represent reality, I've never seen any reason to suspect that they don't follow a consistent and useful cause-and-effect behavior model. A few examples:
* When I look around I perceive physical objects around me which appear to have colors, textures etc. Scientific experiments and tools have demonstrated that there's more to these objects than what I can perceive just by casually looking at them, but nothing about that suggests that my perceptions aren't the systematic result of a useful process for perceiving relevant differences in those physical objects. My computer looks and feels different from my couch, and I have no reason to suspect that those differences in my perceptions do not accurately illustrate differences in the actual materials.
* When I dream or daydream, I can experience things that I wouldn't necessarily be able to experience while I'm awake. However, I appear to be able to consistently identify which of my experiences are dreams and which are not by virtue of a certain quality of awareness which I'll attempt to explain here. When I'm dreaming or daydreaming, everything in the dream is created by my mind, and therefore I have a certain awareness of every dream element that I don't have when I'm awake and perceiving the real world. In the real world, I'm aware only of what I can perceive through observation. Also, my dreams don't seem to be constrained by the same physical limitations that my experiences when awake seem to be. Furthermore, my dreams seem to deal with things that I've experienced or thought about in ways that seem to have a logic all their own, so even within dreams I suspect that my perceptions are still the result of some underlying deterministic mechanism(s).
* While I don't have any experience with hallucinogenic drugs, I do have the consistent experience of not hallucinating while not taking hallucinogenic drugs, so I can confidently deduce that any inconsistencies with those perceptions and an objectively reality could be attributed to those drugs.
And, while it's true that I can't be sure exactly what the underlying causes of my perceptions are... that is to say, I can't be certain I'm not plugged into The Matrix or that I'm just a brain in a vat being fed neural impulses by some system which monitors my own thoughts and generates a complex and consistent illusionary world... in real time... that doesn't really matter. Since I have no way of being more certain about whether or not that I'm observing an objective, unfabricated reality than I currently am, it's epistemologically justifiable to assign that level of confidence as 100% and use it as the standard for which confidence in other propositions can be measured against.
Well, there is a certain problem with this. You can have a lifetime of experiences which indicate, as a pattern, that a certain law of nature exists where in fact that law doesn't exist, or at least doesn't exist in the manner in which your experiences lead you to believe it does. For example, the scientific discovery has been made that water can transition directly from a solid to a gas and vice versa under certain combinations of temperature and pressure. However, prior to that discovery, it's reasonable to assume that most people would live their entire life never seeing such a phase change in water and might reasonably assume that such a phase change cannot happen, and that water must pass through a liquid state when transitioning from a solid to a gas or vice versa. Given our current knowledge and the state of modern science, we can look back and see that such a conclusion would not be true and therefore couldn't be called knowledge. Therefore, we must conclude that the method by which such a person would arrive at that conclusion is not sufficient to determine knowledge, i.e. does not constitute proof.
The problem with basing beliefs on probabilities is that probabilities are assigned based on current knowledge and beliefs. If I flip a coin, I would assign a roughly equal probability that it will land with one side up versus the other. However, an omniscient observer would be able to determine, based on superior knowledge, exactly what the result would be with 100% certainty. So the probability of the outcome is not an inherent and objective property of the event, but rather is as estimate of the likelihood of any given outcome based on currently available limited knowledge.
Essentially, probability is an expression of confidence in a proposition rather than an objective observation... it is the end-product of epistemological study rather than a tool to be used in an epistemological investigation. However, it is worth noting that because we build knowledge upon other knowledge, it may be useful to take an assigned probability of one proposition and use it to help determine justifiable confidence in another proposition (which may be what Hambydammit was trying to say). Exactly how such compound assessments of probability should be handled is, perhaps, a worthy topic for epistemological debate.
I don't think this argument can work Quasar, because in the first few paragraphs you have stated that contiguous observations are sufficient evidence for contiguous observations and not sufficient evidence for knowledge. But here you have moved to say that a lack of observation constitutes a lack of knowledge, this is contrary to your first statement from which it would follow that lack of observation is lack of observation, and nothing more is evident. To make observation and knowledge synonymous is to make an inductive statement that knowledge will always be observed by the knower. (NB you've already contradicted this statement in having said that knowledge must necessarily exist in the absence of contiguous observation.)
So to say, if we render a lack of observation insufficient grounds for knowledge then we have equated having knowledge with contiguous observation but you had earlier said that observation only gives knowledge of itself thus is not sufficient proof of knowledge.
The problem with probability is ascertaining that its only value is to predict outcomes in isolation. That we cling to the notion of isolating lone predicates so defiantly irks me. Probability is an area not a truth value, ie sets are equal to functions not values of functions (even sets of 1) and it's erroneous to conflate f with f(x). The problem of induction stems quite largely from this conflation of terms. From it we get forward moving absolute time - which is false, empty space - which is false also and these are major presuppositions.
If you separate the function from the value of a function the difference between observation and knowledge is perfectly clear and straightforward. The function tends to knowledge, the value of the function is the observation. The prediction regardless of how widely it is validated, is not knowledge it is just valid observation. This is a line that cuts both ways or else we are going to necessarily assign any valid observation a truth value regardless of how ludicrous the mechanism is that the observer (read: possibly theist) comes up with, because we are not assigning the truth value to the mechanism but to the observation. If-Then anything goes.
Therefore any valid belief we have must necessarily be based on probability, not on probabilities - there is a difference - The probabilities only serve validate the belief, but probability is the basis of belief.
Theist badge qualifier : Gnostic/Philosophical Panentheist
www.mathematicianspictures.com
Thank you for the criticism; let me see if I understand your challenge correctly. O = observation, K = knowledge.
O implies O
O does not imply K
not O implies not K
not O implies not O (derived from negation of O implies O, though also apparent through an identity axiom)
And you're claiming that these 2 statements are logically contradictory?
not O implies not K
not O implies not O
If so, I have to strongly disagree.
That said, I'll acknowledge that within the set of logical statements you're listing here, neither O nor not O implies K, which I would agree with using the strictest definition of knowledge. I tried to indicate this in my previous post in the sense that regardless of what I observe, I can never know with certainty the true underlying nature of the thing being observed... all I can know is how it appears to me in my observation.
I think here you're referring to where I propose that, since knowledge in the strictest sense is unattainable through observation, that it's epistemologically justifiable to assign 100% confidence in the proposition that our observations reflect a reality which obeys some set of consistent rules. I want to be clear that I in no way propose that observation should be considered synonymous with knowledge in the strict sense that I used it in my previous post.
That said, if I understand you correctly, you're making the point that my line of reasoning makes the assumption that it's impossible to gain knowledge without making any observation. In other words, I understand you to mean that I don't properly address deductive reasoning like math and logic. But here is the problem with math and logic... the conclusions we draw from them only apply to reality to the extent that the axioms and givens used to derive those conclusions apply to reality. So, you can select any axioms and given statements you want, and derive certain conclusions from them, and claim to have knowledge as a result... and I would agree that you have gained knowledge of a mental model which may or may not correlate to the real world. In order for that to constitute knowledge of objective reality (which is what I was using the term knowledge to mean in my previous post), you have to already have sufficient knowledge of objective reality to know that reality is constrained by the axioms and given statements that your mental model is based on, and I challenge you to provide an explanation for how any perceived knowledge could be epistemologically determined to constitute knowledge of objective reality without the use of any observations.
I have absolutely no idea what you mean by the sentence in parentheses preceded by the nota bene.
You're saying that not O does not imply K, therefore K = O? If so, I'm gonna have to call bullshit on that one. To conclude that K = O, you would need to establish that O implies K (which by negation yields not K implies not O) and K implies O (which by negation yields not O implies not K).
However, in retrospect, I think I may not have been clear enough in my use of the term knowledge in my previous post, so I'll try to give a better explanation of what I mean in the hopes that it will clear up any misunderstandings we may be having. I'm proposing that we know what our observations are, but that we can't know everything about the situations we observe. I'm proposing that it's epistemologically justifiable to assume that, and to act as though, our observations are not random or part of an elaborate deception. I'm proposing that if our observations are consistent and useful that it's epistemologically justifiable to draw certain conclusions from them, but that it's also not epistemologically justifiable to draw certain other conclusions from them. I'm not attempting to provide a complete framework for which conclusions are justifiable and which are not, but to use my example of phase changes in water, I would say that it's justifiable to conclude that water will always transition through a liquid state in situations that are functionally equivalent to those situations in which water has been observed to behave that way, but that it's not justifiable to extrapolate that water will behave that way in all possible situations. This avoids the problem of water skipping the liquid state under certain combinations of temperature and pressure.
No, I am saying these statements are logically contradictory-
O does not imply K
not O implies not K
As you appear to have noted below.
Yes. I am, in the bracketed part. I agree that it is epistemologically justifiable to assign 100% confidence that observations reflect a reality which obeys rules. I disagree when you say that real knowledge must necessarily exist apart from observation. Your statement only implies consistency within observation, that is all that can be inferred from the conditions.
Here I was referring to not O implies not K which implies that observation and knowledge are synonymous and makes an induction from the example you posted.
Exactly. And I'll make the point now that I am not saying that such is impossible, I am only saying that the reasoning you put forth does not provide sufficient conditions for that.
Actually, No. I'm not referring to any of the usual suspects here, surprised?
My point deals strictly with the line of your reasoning which leads to the agreeable conclusion that observation is contiguous and ordered but does not extend outside of that to contiguous and ordered knowledge in the absence of direct observation.
What I am saying is that your line of reasoning does not extend logically to an objective reality. It favourably points to an objective process vis a vis observation, but does not establish that this objectivity is possible to extract from the original observation. It is objective within the observation, take the original observation away and you have no context for this knowledge that you can say is objective.
Ignore it, it was just a passing mention.
No, I am saying that If not O implies not K then O=K because not O implies not O.
Okay, I think that this part may have been too obscure. I was saying as I noted above: If not O implies not K then K=O because not O implies not O. And I am saying we cannot do that because we have already agreed that O does not imply K and obtained that O does not equal K.
I am comfortable with that entirely. I just do not believe it necessarily implies an objectively knowable reality, to me it merely implies a reality in which observations are a reliable source of knowledge about observations and justifies confidence in the said reliability.
Again I agree. But again I think this only ascertains confidence in the conclusions regarding the reliability of observation as observation. It is this part which I emphasised in my last paragraph on probabilities and probability. We can say that Probability is knowledge when it is validated by observed probabilities. We have not implied, however, that the probabilities themselves are knowledge.
I am agreeing wholeheartedly with your logic here, but not with your terms. To borrow your analogy and condense the point - the water states and the situations are not sufficient to be referred to logically as knowledge, they are observations which justify the epistemic confidence in observational order and in being repeatable and widely representable they entail a reliable epistemological justification for the confidence you spoke about, but only in regards to observation. Not in regard to knowledge.
Theist badge qualifier : Gnostic/Philosophical Panentheist
www.mathematicianspictures.com
Then I'm sorry, but your logic is flawed. not O implies not K means that K must imply O (by negation) but makes no requirement about what O must or must not imply. I could make the statement that O implies not K, and it would not be in contradiction with not O implies not K... although, doing so would mean that K must always be false.
I wouldn't say that knowledge gained from observation is not real... just that it's not complete, and that people often draw conclusions from observations which do not constitute knowledge.
Sort of... I don't mean to suggest that the behavior of reality is only consistent while it's being observed. I do make the case that, since our observations don't reveal everything there is to know about the situations we observe, the consistency of our observations can be dependent on some variable(s) which we're unaware of and simply happen(s) to be the same each time we make an observation... and that, as a result of that, it's faulty to claim knowledge that the behaviors we observe will continue to be consistent in any situation.
Then again, your logic is flawed. In order to show equivalence, you have to show that each variable implies the other. Implication in only one direction is not sufficient to show equivalence. Also, not O implies not O doesn't really demonstrate anything... it's just a usage of an identity axiom. Although this:
seems to suggest that you think I'm using the flawed logic in question? In which case, I can only assume that we have some confusion about what we're using K to mean. This would be understandable, since I do suggest that, having made an observation, we know what we observed... however I tried to avoid using the term knowledge to refer to this in my first post to try to prevent such confusion. When I referred to knowledge in my first post, and K in my second, I was referring to complete knowledge of the event being observed. At no point do I try to equate observation to knowledge in that sense.
Fair enough. Then, I'll state that I'm using knowledge to mean true, justified belief about reality. A proposition which is imagined, but which is not supported by any observation whatsoever, may be true by coincidence, but can't really be said to be justified, since there doesn't appear to be any way (again, without observation) to distinguish it from other propositions which are imagined, not supported by any observation whatsoever, and are false by coincidence.
That is the point I'm trying to make, yes. The logical conclusion that we get from extending this line of reasoning is that, no matter how many, and how varied, our observations are, we can never know that we've really understood the true underlying nature of the things being observed... and that we can never know that there's not some condition or variable that, if it were different, would cause our observations to be different.
The point where we disagree seems to be that I think that the observations do apply to an objective reality... just that they do not reveal the complete truth about that reality. Even in the case of The Matrix or the brain in a vat, those are still objective realities which we would be observing, it's just that those objective realities would be very different from what we would expect. It doesn't mean that our observations don't tell us anything about those objective realities, it just means that we don't know the whole story.
I think it would be useful to extend my water example here. The person in the past, who has only observed state changes of water that pass through a liquid state, has observed water changing state from a solid to a liquid to a gas. While he doesn't know that water always passes through a liquid state when transitioning from a solid to a gas, he does know that water is capable of being in any of those three states, and that it's capable of changing state. He may not know why, how, or when it happens, but he knows that it can happen in the right circumstances. With further observations, he might learn that temperature is a determining factor (because he can repeatedly cause the change by adjusting the temperature), but that's not enough to determine that temperature is the only factor. He would be unlikely to discover that pressure is also a factor because, for the purposes of this example, he will die before fluid pressure is discovered. Similarly, today, we know of two factors... temperature and pressure... but we don't know that there isn't a third, undiscovered, factor that we haven't learned to modify yet, and we don't know what temperature and pressure really are although we do know enough to be able to measure them and manipulate them under the circumstances we're familiar with.
We appear to be in agreement here as well. That's why, in my first post, I tried not to use the term knowledge to refer to conclusions drawn from observations, but instead to say that we can determine epistemologically justifiable levels of confidence through observation.
I think you're using the term probability to mean the same thing, though I hesitate to use that term in this way, because to me probability represents a branch of mathematics that deals with hypothetical situations, was invented to determine useful odds for gambling, and appears to be regularly misunderstood and misapplied to try to justify things which it doesn't justify. Specifically, probability is used to justify the belief that a single event can have more than one possible outcome, and that the likelihood of those different outcomes is accurately represented by a probabilistic model... and probability in no way provides any such justification.
You're right, of course, but K v O commutes, no?
Sorry, I didn't mean it that way. By "real knowledge" I am referring to "the gap", ie you are stating that the gap between diverse person's observations is real knowledge even to the person that hasn't observed it. Now I'm not arguing that it isn't 'real knowledge', I think it is, what I am arguing is that your line of reasoning hasn't the conditions to describe it as real knowledge. That is to say, if we both agree that the gap between, say, a scientific observation of water transitioning from solid to gas and a lay observation of water always needing to pass through liquid form, is real knowledge, and I will venture that we do - Then - we agree under different conditions than those which you have stated.
Okay, but even if it's not what you are intending my point is that this is the extent of the assurance that your argument can give us. Reality may or may not behave consistently when we are not observing it, and I would argue that it doesn't BTW, but my contention is that a consistently behaving reality beyond observation is not given by the epistemic justification for trusting observation which you showed.
I need to break this up as it's getting a bit long, and its getting quite late where I am now, I should consider resting. I'll come back to the rest if you won't mind, and thanks for an excellent discussion so far.
Theist badge qualifier : Gnostic/Philosophical Panentheist
www.mathematicianspictures.com
Yep, commutativity applies to conjunction (AND) and disjunction (OR), but not implication.
I'm sorry, but I'm having trouble following your argument here. Can you please illustrate your point with an example?
I see... you are correct, I haven't provided justification for the assumption that the rules that reality follow are consistent when not being observed. As an extension of not being able to know the true nature of reality, we can't know that reality's true nature doesn't allow for parts of reality which are never observed to behave differently than the parts of reality which are observed. If they don't, then I would say that whether or not an event is observed would qualify as the kind of "undiscovered factor" which affects the behavior of events that I've illustrated. In this case, it would also be an undiscoverable factor because we can't gain knowledge of that which we don't observe. However, I would argue that that in no way makes the observations we do make any less useful.
Although, I'm very curious as to how and why you would argue that reality doesn't behave consistently when not being observed. That point of view seems highly counterintuitive to me.
Sure, no problem... take all the time you need. I'm starting to get a bit busy myself, actually, but I'll keep checking for responses here, and I'll try not to keep you waiting too long after your posts. I'll be looking forward to it.
Don't believe anyone , especially yourself , you lie , like me , listen , just listen ? !
G A W E D ? ...... now AWAKEN ! , Why worry said the Buddha ! Oh yeah , the religious walking dead ..... beware of you ! ....
Atheism Books.
Dammit, I had the second half of my reply 2/3rds written last night and then I saved the wrong file. So I have to start again..
Yep, It's not unlikely that I am making a mistake here and if I am I'll blame that the proposition is only negatively defined which is annoying to work with, however; where I am going with it is this -
Because 'O does not imply K' is equivalent to 'K v O' (thus saying we can have Observation OR Knowledge because there is no material implication of Observation in Knowledge) and KvO commutes, we have OvK also (which is to say there is no material implication of Knowledge in Observation.) Which brings us back to where we were - not having observed does not imply not having knowledge. That said, though, I might be stuffing this up, and you're welcome to point me out wherever I am.
I knew this was going to be obscure, sorry. It seems that I did a better job of saying the same thing in the next paragraph anyhow so just ignore this for now:
In addition to saying this, I am also agreeing with you that the rules reality follow within observation have objective application, they are real knowledge, so to speak, and in the previous paragraph what I meant was that our agreement that they are objectively real is on different grounds, grounds we haven't yet established. Aside: I do intend to propose such grounds at some point in this discussion.
What we can know, I would propose, is if they (unobserved parts of reality) have to (behave consistently) or not.
If they necessarily must behave consistently, then we are assured that they do, if on the other hand they do not necessarily have to behave consistently, then we are one down many to go...
I, first and foremostly, reiterate my agreement that this does not make our observations less useful. I am in fact of the opinion that it makes them more useful to a degree which we are probably yet to realise.
What we need to jump into again here is the question of where critical realism greatly differs from direct realism which is the careful distinction between the terms f and f(x), it could be said of a critical realist that they are somewhat a stickler for this distinction. To a critical realist, what direct realism achieves essentially is the statement "f(x) is necessary a valid point on f" this is the realism part, the critical part is then in saying that direct realism cannot claim that set U of observed values of f(x) is equal to f.
To show why using your example take the lay observation of water transiting through liquid state. If we graphed f such that it represented this knowledge we could take the two observed variables temperature and state, and construe the line of correlation over points of f(x) which represent that 'as the temperature rises, the state becomes more fluid'. The lay observer is oblivious of the atmospheric pressure thus it is not represented in their knowledge. We agree however that it is probably still represented in reality.
Now contrast this image with f such that it represents the knowledge construed from the low pressure observation. What we have effectively done by adding another variable to the observation is extended f into an entirely new space. So we say f(x,...) has a representative point where precisely melting temperature equals boiling temperature this is an observed validation. But of f (the abstract representative of knowledge in this analogy) we must say it has an extension in space.
one easy result here is that along the entire curve of f for the lay observation we are carving along a plane in the space of f for the observation of sublimation, this plane probably lies on p=1atm. this is generally speaking, where direct realism / logical positivism begins and ends it's analysis. Clearly this result supports the implication that the lay observer has partial knowledge of the space known to the other observer.
The sticking point for critical realism is that the function is the knowledge. It is, for all intents and purposes the representative form of knowledge. Thus all points observed on f are partial knowledge. the observer of sublimation does not quantitatively have more of the function than the lay observer, each observation is represented fully by a plane within the same space occupied by the function. the difference between the observations is then not a quantity, but an orientation of the observed plane.
To cut a long story short this argument should lead us to consider Topos theory in the sense proposed here:
http://philsci-archive.pitt.edu/archive/00000192/
I would argue that it is not required to, as I mentioned above. The argument that it is not required to is premised on the idea in the last paragraph I wrote of observations being partial truth 'slices' through the phase space of functions of knowledge. Observation is, then, an analogue of orientation.
What we establish in the logic of direct realism does not preclude us each being in different universes, to the extent that even incidental proof of us sharing a universe such as "I am talking to you" is as equally proof of communication across universes as it could be proof of us co-inhabiting a one universe.
And I think that's about enough of the controversial propositions for one night, don't you? I'll sign off there and look forward to your reply.
Theist badge qualifier : Gnostic/Philosophical Panentheist
www.mathematicianspictures.com
Ouch. Sorry about that....
"O does not imply K" doesn't imply any relationship whatsoever between K and O. That is to say, any possible combination of true/false values for K and O will not contradict the statement. "K v O" (as a standalone expression required to be true) requires that at least one of K or O be true, so the 2 statements are not equivalent.
Okay, I think I finally see what you mean by this, and I fully agree with your reasoning and your representative model. I've yet to read the article you link to, but I will.
Ah, yes... and similarly, "I am talking to you" is (subjectively) equally evidence of interacting with another person like the speaker or of interacting with an elaborate deception.
And yes, I'm quite willing to leave these issues alone until after I get to read the rest of the reply you've been working on.