|
Author
|
Topic: Cosmogony, Holography and Causality
|
jasonyoung
Member
Member # 432
|
posted 19. October 2003 21:25
Mr.Langan said: "I don't expect you to know what this means, but I don't expect you to pretend that you do either"
One wonders if gedanken was trying to be ironic in the posts which follow this comment.
Gedanken said: "But of course though precision and imprecision would seem to be opposites or compliments"
Please define precision such that it conforms to a potential and specific state of affairs in the world of perception. When this is done, feel free to concede defeat.
"such categorization in the human is quite apparently an approximate affair with regard to the real-world"
Apparently? It's clearly an approximate affair with respect to some concepts (especially those concepts which concern the emotional and mental states of subjects), but with others it clearly is not, e.g. If it is the case that A, then it is not the case that not-A, where A is a possible state of affairs in the world of perception. That you would deny this is truly baffling.
[ 19. October 2003, 21:36: Message edited by: jasonyoung ]
IP: Logged
|
|
gedanken
Member
Member # 594
|
posted 19. October 2003 23:16
JasonYoung,
The term “precision” was brought up by Mr. Langan, in his comment:
quote:
Gedanken says that raw perception entails maximum uncertainty and unreliability. Since physical reality is perceptual, this amounts to saying that physical reality itself is unreliable, a scientifically meaningless assertion. Scientific observation is anything but imprecise; experiments are generally undertaken with very precise instruments and duly replicated by other experimenters, something which could not happen if gedanken's notions about reality were correct. ...
And I answered that comment with more than one post explaining in some detail what I meant. (One point was that my use of the term “maximum” was a bit of hyperbole, but even that could be supported under certain understandings or views.)
I fully understand that “precision” is typically used with a model of the real world something like the following: The real world has some definite characteristic which we perceive imperfectly. We know we perceive those characteristics imperfectly because if we observe characteristics repeatedly we often find slight differences with regard to our previous memories. This is most notable when a characteristic has a quantifiable aspect. (However quantification, as in assigning numerical value is not directly a “raw perception” but I will get to that in a moment.) So precision is normally referring to that variance of (possibly processed) perception of the real world from the so-called “actual value” or characteristic that we might hold as having some steady value that we feel we come to know as we observe repeatedly with our raw perceptions.
Now sometimes “precision” is used in a strictly numerical context, i.e. with error bars and ranges or tolerances. Often in such circumstances it is thought that one can put absolute bounds on errors. But at other times it is a much more vague concept, referring to degree of correctness in some way but with no particular or well-defined scale or measurement system.
But as I commented at length, even “raw perception” suffers from difficulties caused by the human nature. We realize that our memories fail. So about all that can be completely relied upon in a certain sense is the immediate raw perception. But from experience, we know that even that can be faulty guide to what is happening in the “real world”. The best we can do is to depend upon repetition and careful action in having “raw perceptions” so as to assure by that care and procedure to gain better understanding of the “real world”. We use careful procedure to augment our possibly failing memories, such as recording our perceptions with language immediately as they occur. (A slightly processed version of that could be the sequence I gave several posts above of recording a measurement of a meter.)
quote:
Apparently? It's clearly an approximate affair with respect to some concepts (especially those concepts which concern the emotional and mental states of subjects), but with others it clearly is not, e.g. If it is the case that A, then it is not the case that not-A, where A is a possible state of affairs in the world of perception. That you would deny this is truly baffling.
First, what perception will we talk about that is not a memory at the time we finally get around to discussing it? And is your memory perfect or have you ever known it to fail?
Of course I understand that we use a mental model of a simple logic for classification, wherein we name apparent states of perception such that we expect a crisp delineation of those classifications. So I completely agree that we can have a concept that we are having a particular named state of affairs, such that we think that there is no mater of degree involved. But that is a mental state, that characteristic is more about our thinking processes than about the real world. In other words we appear to be portion of that real world and can treat ourselves, including our perceptions, with treatment we give other aspects of the real world in some manner.
Just I am claiming that in doing so, that we realize that even our perceptions—necessarily taken as memories by the time we are to discuss them—are even themselves not entirely crisp. We can verify that by verifying that our memories are imperfect. Find something you remember, and which you wrote about in great detail. Compare your memory to what you wrote (thus you are not disagreeing with someone else’s perception, but your own record of such). Was your memory always perfect? Did you discover things you did not remember? How does that memory improve as the proximity of events to the present is greater—to what degree? Is there some stepwise degree of time proximity at which memory becomes perfect? If not, now can “raw perception” even be discussed in completely and undauntingly precise terms?
What I am saying is the excluded middle, that A implies not ~A, is a mental concept, a model for treatment of the real world. We develop that model, and find it very useful.
JasonYoung, when you say “It's clearly an approximate affair with respect to some concepts”, where is the dividing line to which you reach those other concepts. How do you determine which of those concepts are “clearly an approximate affair”, as opposed to those other concepts that are not?
That is why I posed the test, to have Mr. Langan speak about a state of affairs (of perception perhaps). And I showed that even that chosen state could have difficulties being completely and absolutely crisply described. Mr. Langan retorted “What is this, gedanken – ‘Fun With Semantics’ time?” (Which was telling, since the mapping of symbology and syntax to semantics was of course precisely the subject.)
Then Mr. Langan went on to describe construction of models, defining terms with respect to the model, formulating a description and a negative description. And of course that is all very useful procedure. In our mental constructions we clearly can imagine a crisp world in which these absolutely crisp delineations have meaning. To create such crisp and absolute models greatly simplifies our task of understanding the world. All I am saying is that it is the model that has the case A that is not the case of ~A. In our thinking processes, we can imagine a world which obeys completely crisp categorization and labeling or naming.
But I’m just claiming that due to our faulty memory, even perception (an aspect of ourselves, necessarily in some manner a part of the real world if “real world” is to have meaning) has matter of degree associated.
Now the term “precise” does have another meaning within our models, such as the real number system. We can model the real world of dimension as though objects fell on positions of the real number line. And thus in our model, we can have the term “precise” refer to a position (in our model) falling exactly in a mathematical sense to a particular real number, in other words we have X=S, where X is real world, and S is specified value, and ‘=’ is mathematical equality. (Of course we have problems making the real world stick to that model!) But I’m certainly not claiming that we don’t make such models, in which we define that such precision and concepts of the excluded middle exist. I never said that such models don’t exist.
One important point is not to conflate the two meanings of “precise”. One meaning is one of degree, the other is a mathematically defined logical statement (defined in our mathematical model). Mr. Langan was clearly using the first meaning in his statement, as science rarely claims a completely accurate numerical equality, and that is left to modeling fields like mathematics. In fact I find it ironic that Mr. Langan should have used that term (wherein science typically determines “precise” with error bars) when the meaning was intended to bring images of the latter definition. But surely Mr. Langan knows of this difference, my surprise is at how pointing this out helps make my concept more understandable rather than implying a rejection.
My point is whether aspects of the real world even actually obey such concepts as the excluded middle, or if it is only our models in which the excluded middle exists. My claim is the latter.
Frankly, I don’t think the things I have described are even in question. What I don’t understand, then, is why there seem to be repeated assertions that characteristics of our models are in fact precise and unerring characteristics of the real world (and here using “precise” in the mathematical sense, as in a model sense of numerical equality.) In other words, I don’t see where anybody has claimed that our perception of the real world is actually the whole story of the real world itself. And if we are all in agreement that perception of the real world is not precise, and it is our models that have the excluded middle, then where are we in disagreement?
(The disagreement seems in conflating the model and the real world. One may argue for such a conflation, but one should not feel baffled that someone else might take issue with that, especially when such argument was not presented in proximate fashion or referred to in the statement of bafflement.)
[ 19. October 2003, 23:48: Message edited by: gedanken ]
IP: Logged
|
|
Rex Kerr
Member
Member # 632
|
posted 19. October 2003 23:31
I wonder whether jason noticed the conversations about a man partially inside and partially outside a room. The point was not that "If A, then it is not the case that not-A" is certain, but rather that neither "A" nor "not-A" could be ascertained with certainty. As such, the certainty of "If A, then not not-A" is merely a curiosity with little bearing on our certainty of any state of affairs that references particulars of the real world (i.e. isn't conditional).
gedankin came up with many examples of how difficult it is to classify raw perception such that it robustly coheres with current perception, and I don't argue with any of them. However, I was trying to refer to the perception-that-I-am-having-now-and-is-ongoing as "raw perception", and what is as certain as anything else is that I-am-having-the-ongoing-perception-I-am-having. Everything else (even very basic things, such as memory) relies upon comparison and classification. And complex schemes often work better there than simple schemes (e.g. study of gravitation gives better results than looking up at the sky imagining chariots carrying things around).
But ultimately, it seems to me that the measure comes back to coherence of raw and immediate perception: I do not see that black dot over there because it is in my blind spot, and the dot really still exists, but I really don't perceive it.
IP: Logged
|
|
gedanken
Member
Member # 594
|
posted 20. October 2003 00:06
Thanks, Rex,
I was going to add another paragraph, but I’ll append it here instead. That was essentially a point that Rex made, but I’ll word it in different terms.
A common model of reality and tolerance or imprecision (in trying to make measurements precise) is that such imprecision of measurement is really a matter of probability. In other words, the model really accurately describes the “real world” in complete detail in some sense, just we are lacking some dimensional values to get completely accurate description along some mathematical axes. In other words the model itself is not questioned, rather the measurement is questioned.
But we see that in case after case, even great precision (taken as a mater of degree) does not mean that the model is completely accurate and that it is the model that should be questioned, rather than the measurement. Relativity is tested in some cases precisely by noting that the content of certain imprecision of planetary motion can be better explained by changing the model from the classical.
But that was an example of exchanging one universal model for another universal model. I am proposing that virtually all such imprecision of measurement is in part due to combinatorial explosion of complicating factors. That the real world does not completely function according to any model. Part of this is that definitions of terms do not even correspond completely to the real world. The models are extremely useful fictions, in abstraction, which can be used for extremely fruitful predictive activity. But the real world is just more complex than that. (And refer to Rex’s comments, and to early pages of this thread.)
We can use the tolerance on measuring a planet’s position in testing relativity theory at relatively low speeds. But the meaning of a planet’s “position” is itself a fuzzy concept. That is not just a question of probability, of the position actually existing just its measurement being in question. Assume a spherical planet. Assume a spherical chicken. Where is the difference?
(And by the way, I agree with Rex on raw perception being most fundamental in some sense. Though of course we must allow for dreaming, hallucination, and other mental states. The “coherence” issue is fundamental as well, as that is how we assure ourselves that we are not diverging into such altered mental states. Things derived from “raw perception” must in some way always remain less fundamental than raw perceptions themselves, even if we judge them to be more reliable in certain ways. We must always remember that the derived concept could be a result of a repeated mistake of analysis, and that it was the raw perceptions that were most “real”. In other words when we change our model of the world, we generally ask that the new model agree with our previous “raw perceptions” as well, though not necessarily our previously generated models.)
--
One thing I have noticed is that philosophy seems upon occasion to not keep up with scientific knowledge. For example I had a physics/philosophy session of some sort in which a philosopher was arguing about “substantial change” occurring in an extended object (e.g. a human) and yet occurring “in an instant”. Now the problem is that nothing can simultaneously occur in an extended object, applying “simultaneously” to all points of the object, and also apply in all reference frames. We know this from relativity. So either the concept of “in an instant” must be in error, or relativity must be in error, or that any object having such a property that relates to its extended form must be in error. Philosophy in this case simply had not been able to deal with the new scientific model of reality, and yet simultaneously deal with those older concepts.
(And note that these are all issues of models in some form. I certainly do not deny the usefulness of logic in discussing models—most certainly not in discussing models that were derived by in part assuming various such logics in the first place.)
I see similar things happening as we learn about the nature of the brain and thought. Mr. Langan’s CTMU paper is even in part an attempt to deal with such questions in some manner—though so far I have apparently disagreed with premises used.
[ 20. October 2003, 00:47: Message edited by: gedanken ]
IP: Logged
|
|
chimp
Member
Member # 333
|
posted 20. October 2003 01:41
quote:
Rex Kerr:
I wonder whether jason noticed the conversations about a man partially inside and partially outside a room. The point was not that "If A, then it is not the case that not-A" is certain, but rather that neither "A" nor "not-A" could be ascertained with certainty. As such, the certainty of "If A, then not not-A" is merely a curiosity with little bearing on our certainty of any state of affairs that references particulars of the real world (i.e. isn't conditional).
The person partially inside "a room" would be evaluated by a many valued logic.
The statement:
"The Earth cannot be rotating in opposite directions simultaneously for any one observer", can only be understood as A or not-A, true or false. It must be an absolute truth.
2-valued logic.
Since this must correspond to what is really observed, it becomes clear that mathematical existence must equal physical existence.
ME = PE
Are you trying to say that logic does not correspond with perceptual reality up to isomorphism?
Gedankin disagrees with logic?
How can Gedankin present a logical argument if he disagrees with it?
How can logic invalidate logic?
[ 20. October 2003, 01:47: Message edited by: Russell E. Rierson ]
IP: Logged
|
|
Rex Kerr
Member
Member # 632
|
posted 20. October 2003 04:21
Maybe the Earth's crust is rotating clockwise and the core is rotating counterclockwise to an observer.
This doesn't look very two-valued to me.
Logic can't sensibly disprove logic, but it can coherently say where it isn't applicable. For example, the logic floating in human-cognitive-analysis-space can be used to evaluate a model wherein logic is the fundamental relationship between abstract entities. If an axiom of the model includes (exists A s.t. A & ~A) then (standard) logic will tell you that logic fails in this model world.
Likewise, logic can tell you whether or not you are justified in believing that logic applies to everything (or something, or nothing) in the real world. (What you are actually doing is building a model for the real world and applying your logic in human-cognitive-analysis-space to that model.)
Subtleties in different layers of model and analysis are easy to overlook. I think that if one looks more closely, logic disproving logic becomes sensible, and links between mathematical and physical existence become rather more tenuous.
Added in edit: I am saying that we cannot find the isomorphism, and thus even if it exists, it's irrelevant as a practical matter.
[ 20. October 2003, 04:29: Message edited by: Rex Kerr ]
IP: Logged
|
|
gedanken
Member
Member # 594
|
posted 20. October 2003 10:47
Russell Rierson said:
quote:
Since this must correspond to what is really observed, it becomes clear that mathematical existence must equal physical existence.
ME = PE
Are you trying to say that logic does not correspond with perceptual reality up to isomorphism?
Gedankin disagrees with logic?
How can Gedankin present a logical argument if he disagrees with it?
How can logic invalidate logic?
Russell brings up an important question—do I disagree with logic? But of course I have repeatedly said that the various logics are very useful. Even here Russell starts with an assumption that “logic” is a single well-defined concept, when we clearly know that there are various “logics”. Incompleteness theorems of mathematics have essentially shown this. (In other words one level of “logic” has shown that there must be multiple “logics” in some sense, even if embedded and nested one outside of the next.) Remember that we define logic, take it as a premise or starting condition of the argument. As such, logics are part of our models.
So I don’t disagree that logics are useful in our models. Just don’t forget where the modeling effort leaves off and the real world begins. If I understand correctly, Mr. Langan suggests a certain meaningful conflation in that regard—but does so in a long argument with many pages. One can simply assert the conflation of model and real world, but I don’t find that to be particularly useful, because to do so immediately invalidates the argument being presented, by the properties of logic that were generally assumed by the presenter of that argument. In other words, Russell, one way to view this is that I am claiming that by your own precepts (including your version of “logic” chosen for your argument) that your statements are internally inconsistent. (That is how “logic” can show a conflict with “logic”—it never invalidated logic, it contested its usefulness in some circumstance in the manner used.)
Of course people may fail to understand that I use various logic level statements all the time, that I use logics as very useful concepts in discussing the real world and the world of concepts. I certainly see nothing wrong with this, as I have presupposed that “logic” (in various forms) is very useful.
Russell, you never answered a question I asked very early: Can you show me a truly “straight line” in the physical world? (That is clearly a “mathematical” concept.)
And yes, I do deny that “it becomes clear that mathematical existence must equal physical existence”. (Quite different from saying that mathematical formulations could approximate physical relationships of existence in some manner, even very useful manner.)
--
Then Russell mentioned that the “person in the room” example could be dealt with by considering “many valued logic”. Of course I suggested the possibility that even continuous valued logics might be considered, in which the number of “many values” of logic corresponds to the real numbers (or say segment [0,1]).
But I contend that the “many valued” logic, though apparently useful, actually complicates the very problem that it attempts to solve. Since these uses of continuous or many valued logics correspond closely to approximation theories, they must necessarily become more complex as they try to model or approximate the real world’s combinatorial explosion of relationships more and more closely. As such, these “many valued” logics fail to be more useful than the original crisp logics, at some ultimate level.
Rather I bring up the continuous and many valued logics as examples of the need to try to solve the problem. And for certain operational situations, like certain “control systems”, the use of such many-valued logics in design of the control system may prove extremely useful because at the level of analysis required the continuous or multi-valued logic simplifies the calculation rather than complicates it. There are cases in which the multi-valued logic is a better model of the real world, with regard to the particular problem being solved in the control system.
In another level, I think that our human thinking process, which by many accounts appears to proceed with something akin to the “neural network” models, actually uses some extremely complex forms of fuzzy logics as part of the thinking process itself. There is of course (once again) no necessity that such fuzzy logics inherently used in the brain should correspond to accurate models of the real world, rather that they are simply “useful” to the human.
[ 20. October 2003, 11:18: Message edited by: gedanken ]
IP: Logged
|
|
chimp
Member
Member # 333
|
posted 20. October 2003 13:49
quote:
Rex Kerr:
Maybe the Earth's crust is rotating clockwise and the core is rotating counterclockwise to an observer.
This doesn't look very two-valued to me.
The core rotates in the same direction as the "crust".
http://www.geo.nsf.gov/geo/adgeo/press/pr9638.htm
quote:
"The inner core rotates in the same direction as the Earth but slightly faster," explains Jim Whitcomb, director of NSF's geophysics program.
Are you are trying to say that logic breaks down at some scale of observation? Do linear, ordered relationships, become chaotic? If a specific n-valued logic only approximates the world, then the "general" logic of probability, agrees exactly with the world. Probability could not exist without a 2-valued logic foundation.
A "mind" operates according to logical rules.
We could say that the reason intelligence exists is simply due to the complex organisation of neurons in our brains. So the question becomes: "What exactly is consciousness?" Does consiousness become the result of an interactive[mathematical?] pattern of communicating neurons...?
What constitutes life? What is the difference between a one celled organism and a complex artificial mechanism, with the same number of interacting parts? If consiousness and sentience is the result of complex interacting structures, then it should be possible to construct an artificial "self aware" computer...
Therefore, a self aware computer program could contain sub-programs within itself, that are also sentient, as well as sub-programs that are not specifically self aware, for example, a background of trees, rocks, houses, etc...
If it is not possible to construct an artificial self aware machine with interacting components, the true mechanism of "self awareness" is a non-algorithmic process?
Interesting...
quote:
Gedankin:
Russell, you never answered a question I asked very early: Can you show me a truly “straight line” in the physical world? (That is clearly a “mathematical” concept.)
This is a question of geometry. Geometry is either Euclidean or non-Euclidean. A or not-A ![[Wink]](wink.gif)
The question "what is space?" is a difficult one. Certainly, for any geometry, whether Euclidean or non Euclidean, a distance interval between two points must obey the "triangle inequality", which states that for any triangle ABC, the distance from A to B to C is never shorter than going directly from A to C.
A triangle on the surface of some geometry, has a rotational invariance, which must hold true for Euclidean and non-Euclidean surfaces:
ABC = BCA = CAB
quote:
Gedankin:
But I contend that the “many valued” logic, though apparently useful, actually complicates the very problem that it attempts to solve. Since these uses of continuous or many valued logics correspond closely to approximation theories, they must necessarily become more complex as they try to model or approximate the real world’s combinatorial explosion of relationships more and more closely. As such, these “many valued” logics fail to be more useful than the original crisp logics, at some ultimate level.
Logic is both general and specific. An approximation is specific and "asymmetrical" A more general statement such as the man is inside the room or outside the room[A or not-A] is a general rule or symmetry.
The general contains the specific.
aleph_0 is the infinity of the natural numbers and it is a general rule which contains the specific elements.
Take a finite quantity away from aleph_0 and it still retains its identity:
aleph_0 - [finite quantity] = aleph_0
Correspondence to the real world, is both general and specific.
[ 20. October 2003, 14:07: Message edited by: Russell E. Rierson ]
IP: Logged
|
|
gedanken
Member
Member # 594
|
posted 20. October 2003 14:42
quote:
This is a question of geometry. Geometry is either Euclidean or non-Euclidean. A or not-A
Correspondence to the real world?
Show me in the real world where to find the mathematically straight line. (I'm not asking where they exist in our models. You keep bringing up the models and claim without justification that they exist in the real world. I'm asking for justification--show me where the straight line exists, for example!)
(Or to put this differently, I have a model in mind of the unicorn. Does the unicorn thus exist in the real world? Of course the straight line has greater relevance to the real world, but that is not the question. We were not asking about relevance, which is a matter of degree anyway.)
As to Rex's point about the core rotating opposite from the outer planet, what is the actual case in the actual Earth at present is not relevant. (That might not be an example of what Rex was pointing out.) What was at issue is whether there might be actual cases of some planet (or the Earth at some time) in which the core was rotating in an opposite direction than the crust. Do you deny the possibility? If so, demonstrate that! (Demonstrate that, down to extremely slow rotational speeds, for all planetary objects!) And if not--where is the definitional clarity that a planet is either rotating one direction, or another, in the real world?
quote:
A more general statement such as the man is inside the room or outside the room[A or not-A] is a general rule or symmetry.
The discussion of the fuzzy logic and case of the person not being either inside nor outside of the room starts on page 4, at my "posted 19. September 2003 13:23". Rather that repeating that argument, I suggest simply rereading the presentations and subsequent discussion already made. That occupied several pages--I see little reason to tread the same ground again.
[ 20. October 2003, 14:58: Message edited by: gedanken ]
IP: Logged
|
|
gedanken
Member
Member # 594
|
posted 20. October 2003 21:34
quote:
A "mind" operates according to logical rules.
I quite disagree with this.
First of all, why do we have courses in logic at college level, if the “mind” operates according to logical rules in the first place?
Of course we can train ourselves in our thinking processes to use logic formalisms, to write and describe according to rules of logic. (In fact that is a very useful thing to do!) But that is a very different statement. (Once again, we propose a set of rules for logic, and test them against real world situations. We determine the usefulness of following those rules, and thus verify a usefulness of a particular “logic”. This is very different from the mind naturally operating according to those rules in the first place.)
From Chruchland, “On the Nature of Explanation” (1)
In abstract: “Neural network models of sensory processing and associative memory provide the resources for a new theory of what explanatory understanding consists in. That theory finds the theoretically important factors to reside not at the level of propositions and the relations between them, but at the level of the activation patterns across large populations of neurons. ...”
This quote mentions a “deducive-nomological” model (D-N), Churchland’s reference [9] (2).
quote:
While much attention has been paid to the logical virtues and vices of this model, relatively little has been paid to its shortcomings when evaluated from a psychological point of view. In fact, the D-N model, and the sentence-crunching conception of cognition it serves, is psychologically unrealistic in several important ways. If someone has just come to understand why a is F, the D-N model requires that we ascribe to that person knowledge of some universally quantified conditional statement having Fx as its consequent, plus knowledge of a series of initial conditions adequate to discharge the conjuncts in the antecedent of that conditional, plus the successful deduction of Fa from this assembled information, or at least the appreciation that the deductive relation exists.
However, while people have an explanatory understanding of much of what goes on around them on a minute-by-minute and even a second-by-second basis, people are decidedly and regularly inarticulate when asked to voice either the general law on which their understanding is presumably based, or the set of initial conditions that tie that law to the explanandum then at issue. Even in simple cases, the premises that people are typically able to supply, when queried, often fall dramatically short of the full requirements of the D-N model. There is no objection, of course, to the idea that people might have large amounts of inarticulable knowledge. What is suspicious here is that the idea that it is both inarticulable and prepositional in character. Furthermore the logical acumen we must ascribe to people on the D-N account is often substantially in excess of what university students with formal training in logic can display.
Further still, the identification of relevant factual premises from a vast store of prior beliefs, and the search for relevant deductive relations, is a process that will in any system consume time, usually a good deal of time. All of this sits poorly with the great speed with which explanatory understanding is commonly achieved. It is often achieved almost instantaneously, as when one understands at a glance why one end of the kitchen is filled with smoke: the toast is burning! Such swiftness is not confined to mundane cases. If one has the relevant conceptual skills, the same speed is also displayed in more esoteric cases, as when one appreciates at at glance why Jupiter is an oblate spheroid: it is a plastic object spinning rapidly; or as when one appreciates at a glance why some red giant close-binary star has the shape of an egg pointed at its more compact blue companion: it is a very large object free-falling in a gravitational field.
At the other end of the spectrum, non-human animals provide a further illustration of these difficulties. Animals too display behavior that indicates the achievement of explanatory understanding, as when a frustrated coyote bites and paws at the leg-trap whose jaws have captured its mate. The coyote understands why its mate cannot leave. Animals too can anticipate elements of the future and understand elements of the present and past, often in some detail. But the assembly of discursive premises and the execution of formal inferences is presumably beyond their capacities, especially at the speeds that faithfulness to their insight and behavior requires.
These particular criticisms of the D-N model are unusual in being empirical and psychological, rather than logical in character. Even so, they are highly general. They will apply to all of the accounts of explanation that require, as the original D-N model requires, extensive prepositional knowledge, relevant retrieval of same, and keen deductive insight. For it is precisely these features that give rise to the difficulties. Is there some alternative way of characterizing the way knowledge is represented in living creatures, and the way it is deployed or accessed in specific cases to provide local explanatory understanding? Yes there is.
(Apologies for long quote, but the material is not available online, and I really didn’t want to simply paraphrase.)
At this point, Churchland goes on to present what he calls the “PDP networks” (parallel distributed processing), which are basically a neural-network based concept. Now my point is not to defend the particulars of the “PDP” concept, or even of any specific neural-network or fuzzy logic based systematization of thinking process. (In fact I’m convinced that Churchland’s thesis is far to oversimplified.) Rather that the argument given against conventional “logic” being used in normal practice by humans is my target.
My major point is that “logical rules” are developed by thinking, over time, and by testing with the real world. They are not inherent properties of the “mind”.
(1) On the Nature of Explanation: A PDP Approach, Paul M. Churchland, Physica D 42 (1990) 281-292, North Holland. Also in Emergent Computation, edited Stephanie Forrest, MIT 1991, North Holland, and apparently also in A Neurocomputational Perspective. MIT Press.
(2) [Churchland’s reference ‘9’] K. Hempel, Studies in the logic of explanation, in: Aspects of Scientific Explanation, ed. K. Hempel (The Free Press, New York, 1965) p. 3.
[ 20. October 2003, 21:46: Message edited by: gedanken ]
IP: Logged
|
|
Claire
Member
Member # 725
|
posted 20. October 2003 22:02
Gedanken,
(words = syntax) why not use post formal operational "thinking" ?(not the word "logic"), remember undefined non linear semantics and excessive power assumed by defined "words" in linear language) However I agree with some of your post. Gedanken quote "...model of the real world..." Model and Real are separate mostly but are linked, this is the interesting part. Russell, be carefull when you suggest "symmetry" is 2 valued but if do you say model AND real is ok then model is only model (not real, see dictionary) and real (very real) is only real. Model can be defined with maths and logic and many other methods too. Real on the other hand could include the same amount of maths and logic that model can, only when we include something else along with it, like another cognitive ability like creative thinking, it, the model or the reality that is real, might have much better coherent qualities that improve our understanding of this real world more cohesively or even the model world. Real might be slightly harder to "real" model though or vice versa. Model could be just as real because it is within the reality that it is trying to model, like the brain, so real and reality as componets of functions of real models might as well be as separate as they are joined. This leaves me with a question, how about finding out "an" answer to the same problem (if there is only one answer or problem) by asking, what is not real or not reality as apposed to what is?
Claire
[ 22. October 2003, 01:11: Message edited by: Claire ]
IP: Logged
|
|
chimp
Member
Member # 333
|
posted 21. October 2003 02:05
quote:
Gedankin:
Correspondence to the real world?
Show me in the real world where to find the mathematically straight line. (I'm not asking where they exist in our models. You keep bringing up the models and claim without justification that they exist in the real world. I'm asking for justification--show me where the straight line exists, for example!)
Concepts like point, straight line, and plane, are the foundations of Euclidean geometry. Euclids 5th postulate does not hold for curved geometries, so there is non Euclidean geometry
A straight line on a curved surface[or curved spacetime] is called a geodesic. An object in "free fall" follows a geodesic. The non-Euclidean example of a Euclidean "straight line". Your question about the straight line appears to be an attempt at diversion? The laws of physics are a subset of the laws of mathematics are they not?
Obviously the solid surface of a planet cannot be rotating in opposite directions simultaneously at one point on the surface of the solid
surface of the planet...?
Now you answer my questions please
Are you trying to say the real world is far too complex for a correspondence to logic?
Are you trying to say that there is absolutely no correspondence?
Are you trying to say that the correspondence is not exact, but only approximate?
I will give you a very specific example of a logical correspondence?:
Take a solid sphere, made of iron, or some other solid material. Consider the sphere as a whole, never mind about the states of the individual quarks and leptons that are the parts of the whole. The ball is rotating on an approximately flat surface.
The ball cannnot be rotating in opposite directions simultaneously from one perspective of observation.
Certain rules must exist in order for the real world to exist
Is Gedankin trying to say that logical rules do not actually exist in the real world?
Is Gedankin sipping tea on Pluto and watching television on Earth simultaneously?
quote:
The discussion of the fuzzy logic and case of the person not being either inside nor outside of the room starts on page 4, at my "posted 19. September 2003 13:23". Rather that repeating that argument, I suggest simply rereading the presentations and subsequent discussion already made. That occupied several pages--I see little reason to tread the same ground again.
The concepts of inside or outside are pretty straight forward. For example put an object in a box, close the lid. The object is 100% inside the box. 2-valued logic. Open the box and begin to pull the object outside the box. It is at first, inside the box 100% and as the clock ticks slowly, the object is a matter of degrees both inside AND outside, which is a many valued logic. Once the object is 100% ...outside the box, it is a 2-valued logic.
quote:
quote:
R.E.R:
A "mind" operates according to logical rules.
Gedankin:
I quite disagree with this.
First of all, why do we have courses in logic at college level, if the “mind” operates according to logical rules in the first place?
Of course we can train ourselves in our thinking processes to use logic formalisms, to write and describe according to rules of logic. (In fact that is a very useful thing to do!) But that is a very different statement. (Once again, we propose a set of rules for logic, and test them against real world situations. We determine the usefulness of following those rules, and thus verify a usefulness of a particular “logic”. This is very different from the mind naturally operating according to those rules in the first place.)
If the mind did not operate according to logical rules, then humans could not function.
Our perceptions of the world must make some kind of sense do they not?
Language itself must be based on logical rules to make sense also.
The mind could not learn college level logic if it did not have the ability to logically process the information.
quote:
Gedankin:
My major point is that “logical rules” are developed by thinking, over time, and by testing with the real world. They are not inherent properties of the “mind”.
Logical rules must exist in order for the world to exist.
Are you saying that the universe does not have rules?
The formal logic that is learned over time exists because the world[universe] is logically consistent.
The general contains the specific .
[ 21. October 2003, 04:29: Message edited by: Russell E. Rierson ]
IP: Logged
|
|
Rex Kerr
Member
Member # 632
|
posted 21. October 2003 04:30
Lots of good points in that last post.
Of course, a solid object cannot (classically) rotate in opposite directions simultaneously. Likewise, an object in free fall will follow a geodesic. An object fully inside a box (classically) is fully inside a box. Except there are no true solids, classical mechanics is only an approximation, nothing is in free fall except in a perfect vacuum with zero radiation, and that's only a classical approximation anyway. Likewise with being inside box.
The issue is that every time one comes up with a classification to use logically, one seems to only approximate the world. In some cases (e.g. the rotation of the Earth), the approximation is extremely good. In some cases (e.g. a single molecule of fluorescin in water ice) it isn't so good.
So when you ask
quote:
Are you trying to say the real world is far too complex for a correspondence to logic?
my answer is, the world is far too complex for us to classify it completely precisely, and without infinite precision, the use of logic only gives an approximation to what will actually happen.
It is not logic's fault, per se; "garbage in, garbage out" as they say. However, it also makes the statement that the world is logical a non sequitur.
quote:
Are you trying to say that there is absolutely no correspondence?
Are you trying to say that the correspondence is not exact, but only approximate?
In some areas there is quite a lot of correspondence! However, there is no support for perfect correspondence. Thus, yes, I am trying to say that the correspondence is only approximate.
One can imagine that if there were a method for perfect classification, perhaps logical rules applied to physical laws governing perfect classifications would exactly correspond to reality. One can also imagine unicorns, or flying through the universe on the crest of a light wave. These are interesting thought experiments, but it doesn't mean that you can fly on the crest of a light wave, that there are unicorns, or that there are Platonic classifications and physical laws in any meaningful sense.
There could be unicorns--but let's find one first. Likewise, let's find the perfect classification/physical laws first, and then believe that they exist.
quote:
If the mind did not operate according to logical rules, then humans could not function.
Our perceptions of the world must make some kind of sense do they not?
I think it is more the case of logic operating according to cognitive rules that allows us to use logic.
Our perceptions must make some kind of sense, but they don't make logical sense in a deep way. (See gedankin's quote from Churchland for arguments against sense being deeply logical.) My impression is that things make sense when they are expected or pattern-matched. Much of this isn't even directly conscious.
There are fascinating experiments where subjects are asked to play "rock, paper, scissors" against the computer for a couple of minutes. The computer has a very simple and bad strategy, but most human players never figure out the logic of their responses or what the computer is doing. Nonetheless, most people beat the computer.
It is indeed the case that the mind could not learn college-level logic if it did not have the ability to logically process information. However, all indications are that logic is a derived scheme based on (less well understood) cognitive primitives, rather than vice versa.
Since cognition is epistemically prior to logic and to (conscious) perception of the world, but perception does not depend on cognitively implemented logic (witness animals that perceive but don't seem to do explicit logic), it does not follow that logical rules must exist for the world to exist. In this particular world, however, it is a very useful approximation to say that they do.
I don't think we can be completely certain that the approximation is perfect.
[ 21. October 2003, 04:33: Message edited by: Rex Kerr ]
IP: Logged
|
|
chimp
Member
Member # 333
|
posted 21. October 2003 04:56
quote:
I think it is more the case of logic operating according to cognitive rules that allows us to use logic.
Logic and cognition are two different forms of the same thing, seems to be what you are saying.
If I take a hammer and accidentally smash my thumb, logic, and cognition, tell me that it will be damaged. Rex and Gedankin will now try to make another semantic readjustment to tell me that the pain I am experiencing does not correspond to what is real?
Logic says that there are consequences for every action.
Rex and Gedankin say that consequences for every action does not correspond exactly to the real world.
I now respectfully ask Rex and Gedankin to define the real world
What is real?
What is not real?
Also, there is the law of conservation of energy, which is as close to an exact correspondence as one can get. Yes or no?
[ 21. October 2003, 05:15: Message edited by: Russell E. Rierson ]
IP: Logged
|
|
gedanken
Member
Member # 594
|
posted 21. October 2003 08:16
Hi, Russell, I'll comment more fully at a later time.
But consider this on "pain". There are people with limbs missing (say due to accident), but they still feel pain in those limbs. Is their pain "real"?
And people get their thumbs (or teeth or any other part) operated on, cut, etc., and don't feel pain. Why, because of modification of their systems by drugs, etc. There is no necessity to a relation of pain stimulus causing events and pain "feeling". But it is a fairly accurate measure of what is important to us. Pain is a very important signal that usually means something significant.
I've pounded a lot of nails. (I've gotten much better at 'hitting the nail on the head' after years and years.) But along the way I've hit my thumb many times, but never seriously. In fact the degree of pain has been quite variable upon striking my thumb, and I can see no damage at present from my past misdeeds.
Russell, did you mean a definitional consideration, that "smashing" is defined as "damage"? If so, you are once again in the world of human models, as opposed to reality. Models are very useful approximations of reality.
Read Rex's post, my position would be similar. How many times have I repeated that these models, and logics are very useful?
But here I really don't understand Russell--how is the association of smashing my thumb and pain a statement of "logic"? One can make a statement (which I did not see), like "if I smash my thumb, I will feel pain". Now that is a statement of logic. (As I have pointed out, the stated rule is not met universally, but is a fairly useful approximation to what actually happens in reality.)
Mr. Langan's use of "logic" went on to very formalized use of mathematical representation. So if you are intending to be using "logic" in that sense, this presentation most certainly did not contain formal language statements of "logic". In fact these vague associations are precisely the kind of fuzzy associations that I am speaking about. The human makes these associations, but they are not strictly matters of formal logic. They are much more accurately modeled with 'neural network' descriptions, than they are modeled with mathematically precise associations of formal logic statements.
I suggest possibly obtaining the Churchland paper, or any of a vast number of papers that discuss body neural function and association mechanisms. We don't need "logic" to have associations of experience and expectation. Churchland in essence points out that even one's dog could make such associations, apparently have such expectations (from learned experience). Does one's dog use "logic" in this association of action and pain?
[ 21. October 2003, 08:47: Message edited by: gedanken ]
IP: Logged
|
|
|
Printer-friendly view of this topic
