An Obsevation I made on Rational Perspectives
I find it hard to believe that in a naturalistic world brand new laws suddenly show up just in time to bring free moral agents into existence.
Labels: emergent laws
posted by Victor Reppert @ 12:11 PM
42 Comments
42 Comments:
Victor
As I've tried to argue in some recent comments, the concept of emergence is of little help to naturalism because of the relationship that must obtain, not just between emergent processes and their emergence bases, but between bodies of rules that meaningfully describe the various layers. Bodies of rules covary dependently across emergence boundaries.
So when C. S. Lewis wrote in Miracles that reasoning must be free from the nexus of blind physical cause and effect, we can transmute his statement to, "The body of rules that meaningfully describe the reasoning process must be independent of the nexus of interdependent rules that describe physical causal processes of various kinds." It may be easier for some of us (some, not all) to see that the rules of inference must be independent of the laws of physics than it is to see that the reasoning process must be independent of neurobiological processes. It is for me, at least.
In this regard, the Stanford online article on panpsychism is interesting, because it proposes that the only alternative to panpsychism is some kind of emergence theory of mind.
What about neutral monism? Or panprotopsychism, Or Spinozistic monism? I think all of these are more plausible than substance dualism.
Darek,
Let's give this another go.
You speak of rules that "meaningfully describe the reasoning process." Can you specify (in detail) one or two such rules for me?
I would like you to show in a specific case why that rule cannot have a physical implementation.
DL
Modus ponens, modus tollens and hypothetical and disjunctive syllogism are usually the ones listed first.
I do not pretend to be either a physicist or a logician. But even my layman's grasp of physics gives me an idea of the type of rules employed in that science to describe physical processes. Likewise, even without training in formal logic I can appreciate in a general way the type of rules that describe human reasoning.
However, to reinforce my basic claim a bit by an indirect appeal to authority, I invite anyone who is interested to google "laws of physics" and then google "rules of inference" and see how close a match they get between the respective lists they generate in either category.
What are you saying Darek, that the laws of physics imply the laws of logic? That they are identical?
BDK
I guess I worded the last part of my post poorly. My point was that what are generally considered the "laws of logic" are quite different than what are generally considered the "rules of inference." I was implying that there is not a good match between the two because they describe different kinds of processes.
To be fair to DL's point, the physical world must be of such a character that the rules of inference allow us to understand it. Let's take an example of modus ponens as follows:
If the sun is shining, it is daytime. The sun is shining, therefore it is daytime.
This concerns physical circumstances and what must or must not be the case in physical terms. But to describe the physical circumstance as rationally understandable is not to describe it as rational in the same way that thought is. We have no reason to believe that the sun employs modus ponens in order to shine in the daytime. Look instead at a thought process:
Mary believes that if the sun is shining, it is daytime. Mary believes that the sun is indeed shining, and therefore that it is daytime.
This makes clear how Mary is relating her beliefs if she is thinking rationally about whether it is daytime. But the process of Mary arriving at a belief about the time of day is a different process than that of the earth turning beneath the sun as the sun generates light. Sure, events must be a certain way in order to be made sense of in the course of thought, but the requirements for being made sense of are not identical to the requriements of thought as it makes sense of events in the world. In order to be made sense of, events must proceed intelligibly; in order to make sense of events, thought must proceed intelligently. Not all that is is intelligible is intelligent.
I hope I don't assume too much if I assert that all of us here believe that "thinking" is a rare and specialized process among a great many non-thinking processes going on around us. If rules of inference peculiarly describe a certain type of thinking--rational thought--then we must place these rules in the category of other special process rules. We cannot say that rational thought is distinguished by being merely intelligible, that is, merely consisting of events that are capable of being understood by rational thought.
My original point is that special process rules or special science rules, if they are truly emergent, covary dependently with fundamental physical laws. But it is difficult, to put it mildly, to imagine rational thought failing to be described (in the peculiar way it is) by the rules of inference if fundamental physical laws happened to be different
Darek,
I hope I don't assume too much if I assert that all of us here believe that "thinking" is a rare and specialized process among a great many non-thinking processes going on around us.
Agreed.
My point was that what are generally considered the "laws of logic" are quite different than what are generally considered the "rules of inference."
We seem to have progressed somewhat. I think we both agree that every system of laws contains laws of logic, or at least some principle of non-contradiction.
You may be correct to say that there's no fundamental law of inference. However, emergent laws do not apply everywhere and at all times.
I think it is misleading to think of the fundamental laws as somehow containing the emergent ones. Rather, we should think of an emergent law as a law that emerges among a population of configurations of matter. We don't think that the laws of chemistry are contained in the fundamental laws of physics, we just think that if you configure matter and energy as molecules, the fundamental laws take on a new, effective character.
Indeed, the fundamental laws are simple and blind to the fact they are acting on molecules and not quarks or electrons. However, this is no barrier to emergence.
So the question isn't whether fundamental laws look like emergent ones. Nor is it a question of the rarity of systems exhibiting emergent laws. The question is whether it is possible to create configurations of matter that exhibit the emergent laws.
Of course, I think it is possible to build a physical, thinking machine because I believe we are examples of such a machine. I also have reason to believe this from computer science.
However, there's no need for you to concede this. If we don't beg the question either way, we would have to say that it is at least possible that one could assemble a configuration of matter and energy as a rational, thinking machine, even if that machine is a zombie.
That being the case, I don't see how your argument has any force to rule out naturalistic minds.
BTW, the naturalistic view is strengthened by the fact that, given physics and non-contradiction, there is an evolutionary advantage to being able to make correct inferences. A configuration that is subject to emergent laws of inference, and can capitalize on those inferences, is selected for.
VIC: "I find it hard to believe that in a naturalistic world brand new laws suddenly show up just in time to bring free moral agents into existence."
Ed's Reply: "Just in time?" Evolutionary time is quite long with twists and turns and about a zillion dead ends along the way. And after all of those twists and turns and dead ends for over a billion years of evolution on this particular planet (the majority of organismal evolution consisting of single-celled organisms reproducing, until the first multi-cellular species finally arose) only a single species has arisen with the capabilities of humanity, though a few large-brained mammalian species like dolphins, elephants, and chimpanzees, rival some of our capabilities of emotion and foresight and sociability in various ways.
DL, you wrote about ‘emergent laws.’ “However, emergent laws do not apply everywhere and at all times.” But isn’t rational inquiry or inference one of those emergent laws that happens everywhere and at all times?
properly basic,
When I say "emergent laws do not apply everywhere and at all times" I mean that they always apply but under certain conditions. For example, the laws of chemistry always apply when the density of nucleons and electrons is in a certain range, and temperatures are lower than, say, 20eV. In places like stars, the laws of chemistry do not hold because the conditions are not right.
Emergent laws that relate to rules of inference would only apply when there are matter-energy configurations capable of applying those rules, i.e., brains. Force the temperature or density too high or too low, or physically disrupt the physical structures of brains, and the emergent rules will no longer apply. Indeed, this is what we observe in humans. Heat the brain and its rules of inference degrade until they cease to apply at all.
DL
I agree that we are making some progress.
Perhaps I can summarize and solidify my previous point by using the phrase, "rules for drawing sound inferences as opposed to fallacious ones," to distinguish these from "rules of intellgibility." As applied to thought, the rules of intelligibility say that a thought sequence cannot both occur and not occur; the rules of inferences indicate how the thought sequence will proceed in order to reach a sound conclusion. Non-contradiction has something to do with both intelligibility and rationality, but not the same thing.
You wrote:
>>If we don't beg the question either way, we would have to say that it is at least possible that one could assemble a configuration of matter and energy as a rational, thinking machine, even if that machine is a zombie.<<
This I don't agree with. "Thinking" and "zombie" are contradictory unless we are talking about unconscious thought. If we are talking about conscious thought of the kind I assume both of us are indulging in right now, we have left zombies behind. A zombie would act as if conscious and rational without actually being either, and the possibility of that I am glad to concede.
The question of whether a machine could consciously reason as you are doing as you read this sentence is far less obvious. One of the ways to answer it is through arguments such my present one concerning rules of inference.
Of course, a bewildering combination of chemical reactions in the brain is a necessary condition of thought as it occurs in human beings. If these reactions are interrupted then inference cannot occur. But my argument is not defeated. In a universe with different physical laws, the human brain as an organ would not be physically possible. Nevertheless, in such a universe thought would be logically possible. But while thought is logically possible in a universe with different physical laws, it is not logically possible that such though could be rational without adhering to the same rules of inference as in ours. It is this distinction which refutes the categorization of rules of inference as emergent from physical laws.
Imagine a table covered with white pebbles. A few of the pebbles are red, arranged in a circle among the larger number of white ones. What if you reached out and mixed the pebbles vigorously, and afterward the same circular pattern of red stones appeared in the very same spot. At that point, you would have to conclude that the circle was not a matter of simple geometric positioning of the pebbles. A red light might be shining down from above, reddening some of the pebbles independent of geometric positioning. Either that, or some other force capable of overriding simple positioning must be at work. It is this kind of situation we are at in terms of the rules of inference versus physical laws, except that we have to mix the "pebbles" with reasoning rather than manually.
Darek,
Taking things one at a time, I would first like to tackle the pebble argument.
When comparing entities across universes, we have to be analogical.
We might say that there are stars in some other universes, and while the stars in the other universes don't use fusion or fuse hydrogen, they do something that gives them some star-like quality. This is possible because stars can be defined in a way that is independent of the detail of the physics behind it. For example, stars might be defined as compact entities that take energy from an attractive force, and radiate it it some other form. Not all universes would support stars, but many universes with radically different laws would support them.
Given this analogy, I could make your pebble argument for stars. That is, I could make the argument that all proper stars possess some stellar quality that shines down from outside of physics.
However, we can see that all I have done here is define some qualities that could be implemented in different ways in universes with different laws (or different ways in the same universe, for that matter). The rules that define "proper stellar behavior" are analogical and universe independent. The implementation of that behavior is universe dependent, but the quality itself is not.
My claim is that rationality is a definition of a set of qualities that could have many implementations (in this universe and others). For example, one plausible definition of rationality would be this:
A rational system is a system that 1) takes facts, and infers new facts according to rules of the universe, and 2) can do the same with counterfactuals.
This definition is a contender because it is a definition we might use to recognize rational agents in an alien universe.
So, that fact that rational systems could exist in multiple universes with different laws of physics doesn't prove anything as far as I can see. We've simply created a definition of a quality that has analogues in other universes.
It cannot be used to argue that rational beings cannot be physical anymore than it can be used to argue that stars aren't physical.
DL wrote:
A rational system is a system that 1) takes facts, and infers new facts according to rules of the universe, and 2) can do the same with counterfactuals.
but what about nature enables us to do 1 and 2? The doing part of rationality seems to sit with us before we experience the rules of the universe. Help me understand where you are going with this. I'm new to this philosophy stuff.
DL
You wrote:
>>We might say that there are stars in some other universes, and while the stars in the other universes don't use fusion or fuse hydrogen, they do something that gives them some star-like quality.<<
Let's keep our camparisons straight. We are talking about the relationship between rules at the fundamental level of physics, rules at higher levels describing emergent physical processes, and finally the rules of inference. In our universe, the laws of astrophysics describe stellar processes. These are emergent from quantum physics (or perhaps string theory/M theory). There are possible universes with different fundamental laws of physics in which different emergent laws of astrophysics describe star-like objects.
Likewise, there are possible universes with different fundamental laws of physics in which thought nevertheless occurs. But the same rules of inference will describe rational thought in any such universe, unlike the case with varying laws of astrophysics. We have an assymmetry between the two examples.
Take your statement:
>>A rational system is a system that 1) takes facts, and infers new facts according to rules of the universe, and 2) can do the same with counterfactuals.<<
To "infer new facts" from previous facts, and to do so soundly, simply cannot be conceived as occuring other than by the rules of inference by which we arrive at sound conclusions about, among other things, possible universes. If different rules of inference generate sound conclusions in different possible universes, then in those universes there is a different definition of a "sound conclusion"--and our conclusions about such universes might not be "sound" in them. No, rules of inference must apply uniformly across possible universes in order for us to make sound judgments by means of those rules. Not so emergent physical laws.
BTW, I will be traveling and may be slow in responding the next few days.
It might be worth reading Frege's classic paper, "Der Gedanke" ("The Thought"). There, he lays out the ways in which the laws of logic differ from the laws of nature. The paper can be found online, I believe, but if not, it's contained in The Frege Reader (edited by Michael Beaney).
Darek,
Let's return to what rationality is/does.
Rationality is that faculty that enables us to explore possible worlds by deducing facts that would be true within those worlds. By comparing these conclusions with experience, we can identify which of the possible worlds we actually reside within.
We can see that this definition of rationality does not depend on any actual laws of any universe. In particular, it does not depend on any actual laws of our universe.
Obviously, a process that is defined to be universe independent will be independent of any universe in which you implement it. The implementation will change, but the process will remain unchanged. Rationality requires only that the universe in which it is implemented be a logical one, which all lawful universes are.
Maybe it will make my point clearer if I describe a situation in which your argument would work.
If materialists were arguing that rationality were defined in terms of physical, extra-logical laws, your pebble argument would work. After all, how can something defined in terms of physical laws in our universe turn out the same in another universe?
However, materialists are not arguing that rationality is defined in terms of physical laws. Rationality is defined in terms of the laws of logic that apply to all universes. It is merely the implementation of rational minds in our universe that is physical-law specific.
Rationality is not a law. Rationality is just a definition or a specification. Minds are systems that implement that specification, and brains are systems that implement that specification physically.
properly basic,
I'm not sure if I understood your question.
Does my last reply to Darek answer your question?
If having a mind required rationality, many of us would be out of our minds.
DL
You wrote:
>>If materialists were arguing that rationality were defined in terms of physical, extra-logical laws, your pebble argument would work.<<
If materialists argue that the rules that meaninfully describe rational thought as a higher level process are emergent the same way as are other special process rules, such as laws of optics, genetics, etc., then they are implying that such rules are dependent upon more fundamental laws of physics. Does dependency equal definition? Not necessarily. But which part of the first sentence of this paragraph, specifically is inaccurate? Do materialists claim that rational thought is a higher level process, yes or no? Do they acknowledge that rules of inference describe the process of rational thought? To the extent that they do so, do they (or would they) compare those rules to other special process rules? If not, why not?
>>Rationality is defined in terms of the laws of logic that apply to all universes.<<
I would ammend that, as previously. Rationality is defined in terms of the rules of inference that apply to rational thought in all universes.
After more reflection on your earlier comments (post before last), I agree that an apparent answer to my challenge would be to make all special process rules strictily definitional. For example, if we propose that the laws of genetics or astrophysics must have precisely the same description and mathematical profile that science now ascribes to them in order to be, in fact, laws of genetics or astrophysics, then the mystery seems to evaporate. The rules of inference must be the same in all possible worlds because such is true of all special process rules.
Ultimately that approach will not work because it leaves no conceptual room for rethinking, correcting and refining scientific laws. The revolution in physics of the early twentieth century rethought and restructured the laws of Newtonian mechanics but somehow retained continuity with them in an important sense (pace Kuhn). Scientists found that the formulations of relativity and quantum mechanics were predictively broader and more accurate and therefore truer. But why does greater predictive power of those rules make them truer? That judgment is the fruit of inference.
Scientists may conceivably change or shift the fundamental rules describing physical reality. Further, I don't think it can be argued that no such shifts could ever affect the conception and construction of secondary physical process rules (the critical question is not whether every imaginable shift would do so, but whether any shift would). An exception must be made for rules of inference because those rules are what would underwrite any such shift in the construal of laws of nature. On what scale do we say that Newtonian mechanics is truer than Aristotlean science, or that relativity is truer than Newtonian mechanics? The scale is the rules of inference. The consensus enjoyed by quantum theory amounts to a collective judgment that it represents the most rational interpretation of the evidence. The same consensus has failed to form in favor of quantum-theory-based efforts to tamper with the rules of inference.
We can only claim to be moving closer to the truth if we have some fixed frame of reference by which to mark our progress, and that is what the rules of inference provide. To illustrate, Einstein could coherently propose flexibility in mass and time measurements and because there was an inflexible constant in relation to which they varied, the speed of light. If the rules of inference employed in making scientific progress were revised, we could no longer meaningfully claim that progress had been made. Scientists are free to revise the rules they use in evaluating physical causes and effects in keeping with the rules of inference, and this goes for fundamental and secondary process rules. But they are not free to revise the rules of inference on the basis of which they make the revisions. Those rules remain constant across shifting scientific frames of reference.
I don't think the laws of logic are fixed.
Also, distributivity of 'and' over 'or' fails in quantum logic. Whether this means we need to overthrow traditional logic, or use quantum logic for the quantum real, I don't care too much. (See here for Stanford Encyc on quantum logic).
I agree with DL largely, but would put it differently (I think it is just semantics). Logic attempts to describe valid inferences (i.e., truth preserving inferences) that don't depend on the content or objects of the particular terms in the propositions. This is what gives logic its generality.
E.g., A-->B, A, therefore B. It doesn't matter what A or B are talking about, their subject matter just doesn't matter. The inference is valid for all A and B.
Though for other rules such as A and (B or C), there is dispute. In quantum logics, that doesn't license the claim (A and B) or (A and C).
Also, there are trivalent logics (with three truth values, to help deal with undecidable propositions).
Talking of logic as if we are talking of how humans in fact think is way too far out there for me. People don't even know how to use conditionals well until they've had years of training in schools.
BDK
The Stanford article contains the following qualification:
"If we put aside scruples about ‘measurement’ as a primitive term in physical theory, and accept a principled distinction between ‘testable’ and non-testable properties, then the fact that L(H) is not Boolean is unremarkable, and carries no implication about logic per se."
This seems to reflect the general lack of consensus that particle physics has disproven the rules of inference. I claim no technical proficiency in the details of quantum theory, but I will happily go out on a limb and claim that physicists are not prone to speaking, writing, or thinking in ways that manifest general skepticism of rational inference--quite the reverse, in fact.
You wrote:
>>Logic attempts to describe valid inferences (i.e., truth preserving inferences) that don't depend on the content or objects of the particular terms in the propositions. This is what gives logic its generality.<<
That's true of most useful rules. The Law of Universal Gravitation says that physical objects are attracted in proportion to their masses, regardless of which physical objects were talking about. If this law applied only, say, to the sun and the moon, it wouldn't be particularly useful. General does not mean vacuous.
Rules that propose generic relations between premises or propositions rather than physical objects are no less real, even though premises are not physical objects.
Ordinary people have no technical knowledge of logic, but that does not prevent them from making rational inferences. People without a deep knowledge of physics (like me) nevertheless attribute a lawlike structure to the world that gives them predictive insight into everyday physical situations.
If the rules of inference that describe rational thought, and that we all depend upon in the course of discussion on this blog, are just the properties of coherence of physical reality then there is a curous implication. "Thought" is a special process, a "higher" process as processes go, can't we all agree? And if thought is a special process, rational thought is still more so. Isn't it a mite curious to claim that the rules that particularly describe the course of rational thought turn out to be nothing more than the coherence properties ("logic") that is at the rock bottom of everthing, even undergirding the fundamental laws of physics? Sort of puts the character of thought in the rarified air of the top of levels of reality and simultanteous on the very ground floor, the foundation. That itself flies in the face of naturalism.
DB:
The law of gravity doesn't come close to being a logical law, as it makes claims about specific properties (e.g., mass). It seems you agree so I'm not sure why you went on about this. Logical rules don't depend on the properties or propositions involved in the inferences.
There is debate about the implications of quantum logic. The quote you pulled is one (slightly odd) way to try to preserve traditional logic by adverting to the difference between 'testable' versus 'nontestable' properties. It is more natural and direct, given the physics, to accept that for quantum systems traditional boolean logic doesn't work. That this is even possible is a problem for those who think logic is insensitive to science.
In QM, everything is twisted all around compared to the ordinary, so it doesn't surprise me at all that the logic developed to deal with medium-sized dry goods doesn't gracefully extend to the quantum realm (without ad-hoc stipulations thrown in about testable and untestable properties).
I think your last paragraph is a potentially interesting attack on a straw man.
I would say that I agree that people can be rational, but that they are often not. Rationality is not a necessary condition for thinking. I can think "If it's raining, there are clouds. There are clouds. Oh my god get your umbrella it's raining." That is a thought. It is an inference. It is a common type of inference (affirming the conseqent). It is a crappy inference to make in deductive logic.
In fact, people are often bad at making good inferences with conditionals, as originally shown with the Wason selection task.
So, be careful to distinguish actual human cognition and inference (which may not be propositoinal in form at all!), from logical rules developed since Aristotle to help people avoid the errors they tend to naturally make, to help formulate rules that will preserve truth as we go from sentence-string to sentence-string. It is common to import the language used to describe this public logic into the head, and then to even say it is a description of how we think (or a presciption of how we should think). This begs lots of questions, in addition to just being wrong, about how we think.
Darek,
Here's how I see the situation.
1) A system of laws connects facts with other facts. In particular, a system of laws tells us what other facts would contradict a given set of facts. For example, a deterministic system tells us what facts about a final state would contradict facts about the initial state.
2) Inference is just the process by which we determine new facts from existing facts within the context of a set of rules. It is the elimination of contradictory possibilities.
3) Inference is a specification for a process that is executed by rational minds.
4) Human minds are complex physical systems.
5) Inference is a "higher-level process" just because it is a process that requires complex configurations. It is higher-level in the same way that organic chemistry is higher-level than fundamental particle physics. That is, it is a higher-level physical process.
6) Inference is low-level with respect to mental activity. That is, it is a low-level mental process.
7) There's no conflict in seeing inference as high-level physically and low-level mentally.
Personally, I find this whole picture to be consistent and aesthetically pleasing.
You say:
Isn't it a mite curious to claim that the rules that particularly describe the course of rational thought turn out to be nothing more than the coherence properties ("logic") that is at the rock bottom of everthing, even undergirding the fundamental laws of physics?
I think this contradicts my (7). Rational inference examines the lowest levels of reality. Rational inference does not undergird the rock bottom of everything. It's not as though the physics or the principles of non-contradiction within our universe rely upon inference. Our knowledge of those things relies upon inference, and that's a whole different kettle of fish.
As for quantum logic... Quantum mechanics shows us that it is possible to have systems of laws that contain notions of contradiction without being classically logical. It has no effect on the specification for what an inference is/does, but it does change the mechanics of making an inference because the logic has changed.
I think Moreland’s or someone else's thought applies here: that if you can't know some things without knowing why you know them--if you don't have some things in place to begin with--you can't know anything at all. You can't even begin the task of discovery. Aristotle said that some things can't be proved, but without them you can't prove anything.
DL
I think you have an obvious conflict between 4) and 6). It is a measure of the physicality of a process that the rules that meaningfully describe it are physical laws (either fundamental laws of physics or higher level laws dependent on the fundamental laws of physics). A process that is meaningfully described by rules that are not physical cannot be claimed as a wholly physical process. The goal of naturalism is to show that the mental is a subdivision of the physical; if there is a fundamental distinction between the two, naturalism as commonly conceived is defeated.
As for quantum mechanics, isn't it fair to ask what rules of inference were used to establish the validity of quantum theory? Classical rules of inference or quantum rules of inference? Whatever rules of inference were used to establish the likelihood of quantum theory are the rules we are stuck with. Paradoxes that emerge in the theory mark the frontier of intelligibility but not the refutation of the very rules of inference that led to the theory's adoption.
Finally, I have thought of a way to illustrate the difference between mere intelligibility and rationality. Look at the following syllogism:
All thought sequences that occur purely according to suggestive association are not described by the rules of inference. Bob's thoughts follow each other purely by suggestive association, therefore Bob's thoughts are not described by the rules of inference.
Now, assuming my major premise is true and further assuming that Bob were someone daydreaming his way toward slumber in accordance with the minor premise, the conclusion seems inevitable. But in one sense that would mean that Bob's thoughts are indeed described by the rules of inference, since the syllogism is a description of Bob's thoughts and it follows those rules. The conclusion would therefore be false if it were true. Is this a hopless paradox? To me, it obviously is not. Bob's thought processes are merely intelligible; they are capable of being thoughtfully analyzed according to the rules of inference. But the actual sequence of Bob's thoughts themselves is not described by the rules of inference.
As for quantum mechanics, isn't it fair to ask what rules of inference were used to establish the validity of quantum theory?
Luckily it was science, so experiments established the validity of the mathematical framework.
Darek,
It is a measure of the physicality of a process that the rules that meaningfully describe it are physical laws (either fundamental laws of physics or higher level laws dependent on the fundamental laws of physics).
I think this is incorrect. "Meaningfully describe" is a code word for the description of a thing's higher function. Something can have a higher function without contradicting its lower mechanism.
If I build a computer circuit out of logic chips, it does not cease to be described by atomic physics nor does it cease to be described by the rules of digital circuits. It's easier and more convenient to describe its operation in terms of what it does or in terms of high-level concepts, but those descriptions are not mutually exclusive of the low-level physics.
As for quantum mechanics, isn't it fair to ask what rules of inference were used to establish the validity of quantum theory?
The general process of inference has never changed. What happened was that certain rules used in classical inferences and thought to be general were found not to be so general. We once thought that measuring AB was always the same as measuring BA, and now we know that this is false.
However, QM doesn't alter the general process of inference. It just changes the rules we believed held in all physical systems. Classical rules were found to be inconsistent with experience, and only a subset of those rules remained valid.
All thought sequences that occur purely according to suggestive association are not described by the rules of inference. Bob's thoughts follow each other purely by suggestive association, therefore Bob's thoughts are not described by the rules of inference.
I think that the premise here fails to hold, and it fails for the same reason that your first argument fails. Suggestive association could be used to make inferences. Not all suggestive associations are valid inferences, but that's not the issue.
We seem to be backtracking somewhat.
Earlier you wrote:
After more reflection on your earlier comments (post before last), I agree that an apparent answer to my challenge would be to make all special process rules strictily definitional. For example, if we propose that the laws of genetics or astrophysics must have precisely the same description and mathematical profile that science now ascribes to them in order to be, in fact, laws of genetics or astrophysics, then the mystery seems to evaporate. The rules of inference must be the same in all possible worlds because such is true of all special process rules.
If inference is a "special process rule", it's clear that some physical systems will not implement the rule, but that some may.
It really doesn't matter what the low-level physics is as long as it permits the implementation of machines that meet the special process rules. There's no conflict between the high-level and low-level descriptions as I explained in my prior comment.
DL
>>I think this is incorrect. "Meaningfully describe" is a code word for the description of a thing's higher function. Something can have a higher function without contradicting its lower mechanism.<<
"Meaningfully describe" can hardly be a code word for "describe." The description involved is non-trivial. This is straightforward.
First, are you saying that the "higher function" is other than a physical function?
Second, are you saying that the higher function is independent of the lower mechanism?
If your answer to both of the above is no, how does your objection make sense?
>>I think that the premise here fails to hold, and it fails for the same reason that your first argument fails. Suggestive association could be used to make inferences.<<
I disagree that the purely associative can be inferential, but I my example can easily be reworked. I can begin, "All thought sequences that are purely random . . .," since the rational "therefore" required by inference simply cannot be random with respect to the antecedent and conclusion. Or even, "All thought sequences in which no conclusion is reached . . ." My point is simply that the susceptibility of a thought sequence to analysis by means of the rules of inference is not the same thing as the rationality of the sequence. Do you fail to grasp the distinction?
>>It really doesn't matter what the low-level physics is as long as it permits the implementation of machines that meet the special process rules.<<
Implementation is a top-down conception. Remember, we are talking about emergence. In naturalism, all special process rules emerge from underlying rules. There could be other rules of physics that would still be, well, rules of physics. There could be different rules of chemistry that would still be rules of chemistry, but will always be emergent from physics; different rules of genetics that would still be rules of genetics; and so on. The rules at various emergence levels will be dependent. Even conventional rules like those of chess and baseball, given naturalism, are emergent from and dependent upon more fundamental rules. In this naturalistic picture, all rules that are genuinely descriptive are interlocked in such a way that none of them can remain stable given every conceivable shift in the others. This is the basis of my argument and I have seen no refutation of it from you.
Darek,
First, are you saying that the "higher function" is other than a physical function?
Yes. Higher function has abstracted away the physical implementation.
Second, are you saying that the higher function is independent of the lower mechanism?
Yes. I can make a calculator out of vacuum tubes, relays, gears, silicon integrated circuits...
The specification for "calculator" transcends physics, and we can imagine calculators in other universes where the physics is different.
My point is simply that the susceptibility of a thought sequence to analysis by means of the rules of inference is not the same thing as the rationality of the sequence.
Sorry, I missed what you were trying to say last time. This sounds reasonable. How is this distinction relevant to our discussion?
Implementation is a top-down conception.
I'm going to continue to use the term implementation, despite its typical use in top-down situations. I don't have a better word for "kind of instantiation," but I'm open to suggestions. :)
Remember, we are talking about emergence. In naturalism, all special process rules emerge from underlying rules.
No, I disagree that naturalism says this.
Let's go back to my calculator example. It is the implementation of the calculator that is emergent, not the calculator itself. The concept of a calculator abstracts away the physical details leaving a pattern that can be matched by analogy.
A Platonic realist (not that I am one!) would say that the calculator abstraction existed timelessly, in parallel with all possible universes. He would say that, in some universes, particulars of the calculator form (implementations) emerge.
In this naturalistic picture, all rules that are genuinely descriptive are interlocked in such a way that none of them can remain stable given every conceivable shift in the others. This is the basis of my argument and I have seen no refutation of it from you.
But the rules are only interlocked for particular implementations of a specification, not for the specification itself.
As soon as you abstract away from the physical implementation, you create a specification that is independent of implementation.
A mind is defined abstractly, not by what any particular implementation does at a low level.
So I think your argument is a sort of straw man. Naturalists do not generally argue that minds (say, in their Platonic form) emerge but rather that implementations of minds emerge. There's a difference.
To the extent that an atom is defined abstractly (such that another universe could have atoms), it would be wrong to say that atoms emerge in the physics of our universe, and it would be better to say that implementations of atoms emerge in our universe. That is, if we apply your argument to atoms, even atoms would not be physical.
DL
>>"My point is simply that the susceptibility of a thought sequence to analysis by means of the rules of inference is not the same thing as the rationality of the sequence."
Sorry, I missed what you were trying to say last time. This sounds reasonable. How is this distinction relevant to our discussion?<<
I was attempting to revisit briefly, and clarify, the issue that had occupied quite a bit of time previously, that is, the difference between intelligibility and intelligence. I'm not sure whether that issue is settled or not. Moving right along . . .
Am I to understand from your last post that you repudiate physicalism while attempting to retain naturalism? That does cut down the room for discussion.
>>To the extent that an atom is defined abstractly (such that another universe could have atoms), it would be wrong to say that atoms emerge in the physics of our universe, and it would be better to say that implementations of atoms emerge in our universe.<<
"Atoms" do not emerge in the physics of our universe? Just implementations of atoms? This sounds like Platonism, to the extent that I can make it out. Suppose we define atoms as actual objects existing at real places and times. Do those actual atoms actually emerge? And are those actual emergent atoms described by actual laws that emerge as the actual atoms themselves emerge?
Even if you don't think that laws of nature emerge in layers, that is I believe an increasingly popular concept in popular naturalism.
Compare the laws of nature around us to a landscape. We see a particular landscape, but we can imagine a different landscape of laws. As long as those imagined laws operated with discernible regularity, we could still make sense of them by means of the rules of inference. But if the rules of inference were themselves part of the landscape, presumably they could be different and we could still make sense of them. That would mean that a logical fallacy could be imagined as a rational course of thought if it were part of a different landscape of laws of nature, but it can't be. A landscape of laws that would make a logical fallacy rational would not be the kind we could make sense of--would not be a landscape of laws at all.
One way or another, naturalism is committed to a single landscape to which everything real belongs.
Now, maybe this distinction between the character of laws of nature and the character of rules of inference can be accounted for by some strange twist of naturalism. I remain skeptical. But to say that naturalism can somehow account for the distinction is different than saying the distinction does not exist. If you are arguing the former rather than the latter, then maybe we can more profitably continue the discussion under another post of Victor's, since this thread is getting a bit long in the tooth.
Darek,
Am I to understand from your last post that you repudiate physicalism while attempting to retain naturalism? That does cut down the room for discussion.
I'm not sure this is relevant because physicalism accepts that there are computational laws. However, just FYI, my personal naturalism isn't physicalism. IMO, a naturalist is someone who rejects non-predictive explanations in any field of inquiry.
Suppose we define atoms as actual objects existing at real places and times. Do those actual atoms actually emerge? And are those actual emergent atoms described by actual laws that emerge as the actual atoms themselves emerge?
Yes. However, when you talk about inference as something independent of physics, you are no longer talking about implementations of inference. You are integrating out the implementation to create an implementation-independent definition. You are talking about an abstraction of inference, not actual inferences as they happen here in our actual brains.
One way or another, naturalism is committed to a single landscape to which everything real belongs.
Physicalism is committed to a single landscape to which every real implementation belongs. There's a difference. It does not commit itself to saying that every possible abstraction physically exists, but only that every abstraction thought of has a physical implementation.
Now, maybe this distinction between the character of laws of nature and the character of rules of inference can be accounted for by some strange twist of naturalism. I remain skeptical. But to say that naturalism can somehow account for the distinction is different than saying the distinction does not exist.
There's no twist. The rules of inference are rules that apply to abstractions. The rules of physics apply to particulars. There's your distinction. If thought-of abstractions have physical implementations, problem solved.
DL
>>IMO, a naturalist is someone who rejects non-predictive explanations in any field of inquiry.<<
Let's take the proposition, "Non-predictive explanations are less likely to be true than predictive ones." Is the proposition trivial, or does it inform? Suppose that it is informative. I don't see how it can be falsified by the test of predictability, and if it cannot be so falsified then it cannot be so verified. On the other hand, if it is not informative why is it valuable as a principle? It seems to me that it is an inference from the assumed intelligible character of the physical world. It a non-predictive explanation that apparently is not rejected by naturalists.
>>It does not commit itself to saying that every possible abstraction physically exists, but only that every abstraction thought of has a physical implementation.<<
What is the a difference between an abstraction and a “possible abstraction.” Is a possible universe with different laws of nature an abstraction that is thought of? It would seem hard to deny it. But does such a universe have a physical implementation? I doubt seriously that it must.
At least you seem to distinguish between abstractions and physical objects/states. Well, the rules of inference are rules for manipulating and relating abstractions. Computers manipulate representations of abstractions according to the rules of electromagnetism, but it is we who invest the representations with meanings that fall under the rules of inference. Just like the computer manipulates representations of clouds and rain according to the rules of electromagnetism, but it is we who see the images on screen as standing for clouds and rain subject to the laws of gravity, thermodynamics, etc. The computer does not implement clouds and rain.
Take the word “sphere.” That word does not implement a sphere, it represents it. An implementation of a sphere must be spherical; the word “sphere” is not. Take the term, “a perfect sphere.” The term represents an abstraction, but the abstraction has no physical implementation (nevertheless we can manipulate that abstraction in reason).
Representation does not necessitate implementation. Therefore, for a computer to represent the rules of inference does not necessitate that it implement them. In fact, because rules of inference are rules for relating abstractions, only a system that manipulates abstractions-a mind-can implement them.
Darek,
Let's take the proposition, "Non-predictive explanations are less likely to be true than predictive ones." Is the proposition trivial, or does it inform?
Sorry, but I must have left you with the wrong impression. This has nothing to do with whether non-predictive models are likely to be correct, but whether they are explanatory at all.
A non-predictive model can do no better than restate one's observations or mental experiences. That's why such models cannot explain, by my definition. If you consider restatement to be explanatory, then every experience has a trivial explanation in the form of a restatement of the experience. That's what my kind of naturalist rejects. The naturalist demands that every explanation make interpolations or extrapolations beyond what is being explained.
Note that the naturalist doesn't say that everything has a predictive explanation. That's too strong a statement. It's possible that some things don't have predictive explanations, but then such things are fundamentally inexplicable. IOW, there's no such thing as a supernatural explanation. That would be an oxymoron.
What is the a difference between an abstraction and a "possible abstraction." Is a possible universe with different laws of nature an abstraction that is thought of? It would seem hard to deny it. But does such a universe have a physical implementation? I doubt seriously that it must.
This is a good illustration of what I am talking about. We must keep track of the levels of abstraction.
I'll see if I can label them:
Layer A0: actual, specific events.
Layer A1: the class of actual events (including events we are not yet aware of)
Layer A2: events, actual and otherwise
Layer A3: actual, specific laws describing layer A2
Layer A4: the class of actual laws (including those not yet discovered)
Layer A5: the class of laws, actual and otherwise
Each layer of abstraction can only be maintained by presuming we will know a member of the abstract class when we see it. For example, we can talk about Layer A4 because we have some confidence that we could recognize a member of the class of actual laws if we saw such a member.
Your specific question is about Layer A5. It has members (possible constellations of physical laws) that almost certainly do not have physical implementation. I think this was your point.
Okay, back to your central argument...
Let's look at abstractions about minds.
Layer M0: actual, specific minds
Layer M1: the class of actual minds
Layer M2: the class of minds, actual and otherwise
Now, in order to create abstraction M2, we are saying we would recognize a mind if we saw one. This abstraction applies to counterfactuals, not just actual reality. I would recognize a mind in my cat, if my cat had a mind. Similarly, I would recognize a mind in another universe, if the universe existed.
Layers An and Mn are basically orthogonal. Saying that M2 carries across universes doesn't say anything about whether the actual minds we know of are physical.
To prove this, I can replace the word "mind" with "plasma" in hierarchy M. Then I can imagine plasmas that do not exist, but that I would recognize if they did. I can imagine plasmas that exist in other universes where electromagnetism doesn't exist, but where the alien phenomena are analogous to plasma by having some analogue of ionization and temperature. Yet, if the ability of an abstraction to span universes were a sign that the abstracted thing were non-physical, we would have to say that plasmas were non-physical, which makes no sense.
The reason that class of possible minds (M2) spans universes is not that members of M2 are non-physical. The reason class M2 spans universes is that the test for membership in class M2 does not depend on specific laws of physics. M2 is defined in a way that is physics independent in the same way that a plasma is defined in a physics-independent way.
Plasmas are actually a great example. Plasmas were originally defined as they apply to electromagnetism. An ionic plasma is a high temperature state where atomic bound states become unbound and form a soup with force-carrying photons. However, we now study plasmas made up of quarks and gluons. Quark-gluon plasmas feature a dense, high energy state of formerly bound quarks in a soup with force-carrying gluons. The forces are not electromagnetic, but strong nuclear - a different physics altogether. Of course, we can similarly imagine such plasmas in alien universes where all the laws are different. This does not make ionic plasmas non-physical.
DL
>>The naturalist demands that every explanation make interpolations or extrapolations beyond what is being explained.<<
Alright, how about this proposition: "Every informative explanation makes interpolations or extrapolations beyond what is being explained." Is that proposition subject to any kind of predictive test? Or is it fundamentally inexplicable?
>>Yet, if the ability of an abstraction to span universes were a sign that the abstracted thing were non-physical, we would have to say that plasmas were non-physical, which makes no sense.<<
An instance of reasoning concerning plasmas is not identifiable with plasmas. It is this instance of reasoning concerning plasmas that ranges over possible universes, no plasmas themselves.
>>Of course, we can similarly imagine such plasmas in alien universes where all the laws on different.<<
I don't think we can imagine plasmas governed by the same rules in a universe with different laws of physics. Every rule-oriented description of plasmas will invoke terms such as "phase of matter" or "energy" or "particles" or the like. These terms can only be understood in the context of physical laws describing them. If the context of physical laws changes, the meanings of these terms shifts with them. So we do not have an equivalence to the rules of inference.
Some further thoughts . . .
I know what an atom is to the extent that I know what it is physically, that is, what physical relations it enters into, how it is quanitified versus other physical objects and states. The less I know what an atom is in physical terms--the less specifically I can place it in its causal context of physical laws--the less I know what it is, period.
Rules of inference, by contrast, relate premises to conclusions. But I have no idea what a premise is in terms of physical quantities and properties, and trying to identify a premise in that way seems absurd. I don't know how to situate a premise or conclusion into a context of physical laws the way we do with neutrons and plasma bodies, quarks and comets. I doubt that anyone else can either.
Given a single naturalistic context of reality, that should mean that our knowledge of premises and conclusions and their interrelationships is extremely vague, if we have knowledge of them at all. But in fact we know them well. We have trouble placing premises into a physical context, not because we have such a poor idea of what they are, but because we have such a good one. It is because we know premises so well, and can relate them to one another with some skill, that we can use them to understand physical events and states.
Yet another way to look at it. The more we know about laws of nature at lower levels, the more we know about upper levels. The more we know about the laws of particle physics, the more we know about the laws of chemistry. The more we know about the laws of chemistry, the more we know about the laws of genetics. Knowing more about the laws of physics, however, does not inform us better about the rules of inference, because our knowledge of the laws of physics depends upon our use of the rules of inference. Even strange twists on the rules of inference can only be recommended on the basis of those same rules.
Darek,
Alright, how about this proposition: "Every informative explanation makes interpolations or extrapolations beyond what is being explained." Is that proposition subject to any kind of predictive test? Or is it fundamentally inexplicable?
It's not subject to any test because it is a definition. This is like asking whether "Every even number is an integer multiple of two" is subject to verification. It isn't because it is a definition.
The issue here is how to define the distinction between something that is explanatory and something that is not explanatory. If one values trivial explanations which are mere restatements of one's observations and experiences, then supernaturalism works just fine. On the other hand, if one considers mere restatements to be non-explanatory, one is forced into naturalism (as I have defined it).
It is this instance of reasoning concerning plasmas that ranges over possible universes, not plasmas themselves.
In a sense this is true. An abstraction is a class with a rule for membership. An abstraction requires a mind. Plasma is an abstraction, of which there are particular physical instances in this universe and others. Therefore, the rule for admission to an abstract class may be independent of the physical details of any particular member.
However, that doesn't mean that all minds are non-physical. That claim has nothing to do with this. Obviously, I am using a mind to create a rule for set membership which does not depend on physical details of its members. However, this does not imply that the mind performing the abstraction is non-physical. You haven't proven that physical abstractors could not create abstractions about other universes (or counterfactual ones).
Just to keep score, your prior claim was that the abstraction for "mind" was portable across physics, and that that meant that particular minds could not have physical implementations. The plasma counterexample demonstrates that your original argument does not work because even if the abstraction for plasma is portable across universes, there are still physical instances of plasmas in these universes. Likewise, even if mind is a portable abstraction, that does not prohibit there being physical implementations of minds in those universes.
Given a single naturalistic context of reality, that should mean that our knowledge of premises and conclusions and their interrelationships is extremely vague, if we have knowledge of them at all. But in fact we know them well. We have trouble placing premises into a physical context, not because we have such a poor idea of what they are, but because we have such a good one.
I think you can only look at the world in this way if we forget about human history.
Before we knew that animals were made of cells and molecules, we were intimately familiar with them. For all we knew, animals were animated magically. We were completely unable to isolate and bottle Essence of Tiger. We could not even imagine mechanisms that would animate tigers. Now we can see how tigers work as physical machines, even if we cannot yet explain every detail. Thanks to science, we no longer believe that tigers have non-physical components.
This seems like a perfect analogy.
1) Tigers are more complex than mere chemicals, as are brains.
2) We were initially very familiar with tigers without understanding how they reduce. The same is true of brains.
3) Over time, we discovered physical mechanisms that partially explain bio-mechanical tigers. Similarly, we have discovered many physical mechanisms that partially explain mental function, including memory, computation, recognition and abstraction.
One could claim that there are still things about mental function that have not been explained physically. Sure, but that's a supernaturalism of the gaps argument (which I don't think you're making).
You are trying to improve on this argument by showing that our pattern of discovery of physical mental function is without parallel. Unfortunately, the tiger analogy shows that this is not the case.
Even if it were the case, you would not have a valid inference to non-physicality merely by showing that the mind was an outlier. Rather you would have an anomaly. To make an inference in favor of non-physicality you would have to show that non-physical systems are typically identified by a pattern of discovery like that found in neuroscience. Of course, like all such arguments for the supernatural, that won't work because no non-physical systems have ever been discovered.
DL
>>It's not subject to any test because it is a definition.<<
Definitions can be informative and explanatory. I think that you provide a definition of naturalism that attempts to explain what an explanation ought to do. So I asked whether your explanation of explanations succeeds or fails on its own terms, but I no longer expect an answer.
>>Just to keep score, your prior claim was that the abstraction for "mind" was portable across physics, and that that meant that particular minds could not have physical implementations. The plasma counterexample demonstrates that your original argument does not work because even if the abstraction for plasma is portable across universes, there are still physical instances of plasmas in these universes.<<
Actually, my claim was that the rules of inference that describe rational thought cannot be imagined to shift with any imaginable shifts in the fundamental laws of physics. Since this cannot be said of dependent bodies of rules such as the laws of optics, thermodynamics, chemistry, etc., then rules of inference cannot be classed with these. Therefore, whatever else may be said about minds, when minds engage in rational thought they are performing an activity described by rules that are independent of the interconnected nexus of physical laws. To the extent that popular naturalism, at least, sees all reality as occurring within this nexus of physical laws, my line of argument undermines naturalism as commonly understood.
Your example from plasma does not refute my argument, for the reason I gave in my last post and which you failed to address.
Is this disconnect merely an anomaly? To the extent that it lies at the heart of our intellectual life I don't think it can be so minimized. In addition, it is not ad hoc with respect to theism. That is, it suggests that intelligence is a fundamental feature of reality rather than a by-product of blind physical forces, and does so independently of the usual reasons for belief in God.
Let's say we assume moral realism. If there are objective moral laws, it is difficult to see how these laws could affect purely physical systems. Some naturalists, such as Keith Augustine over at the Sec Web, reject moral realism for precisely that reason. But not all naturalists find it easy to dismiss all morality as subjective preference. So we have the live possibility of a second anomaly for naturalism concerning moral realism, as for example Jeff Lowder (I believe it was) frankly acknowledged in a paper some years back.
The question of if or how God "causes" rules of inference and morality is to me less interesting than the question of how these rules could affect us, even imperfectly, in the absence of God--a governing Mind that conditions all of reality.
The disconnects between rules of inference and morality on one hand and laws of nature on the other are not the only reasons for belief in God, but they are among those reasons for many, including myself.
Darek,
Definitions can be informative and explanatory.
No, they are not explanatory. Definitions are part of every model, but it is not the definition itself that does the explaining. (Not unless you say that the theorems of a definition are explained by the definition.)
If I define a bunny as that small, furry, mammal with long ears, what has been explained? Can I explain a thing merely by naming it? Or merely by naming its explanation?
Therefore, whatever else may be said about minds, when minds engage in rational thought they are performing an activity described by rules that are independent of the interconnected nexus of physical laws.
We can define rational inference for any system of laws, as long as that system of laws is logical. That means that the definition of inference is physics-independent. You cannot turn around and declare that pre-defined independence to be a discovery about the world.
You claim that I did not address a point you made in a previous comment. I assume it is this:
Every rule-oriented description of plasmas will invoke terms such as "phase of matter" or "energy" or "particles" or the like. These terms can only be understood in the context of physical laws describing them. If the context of physical laws changes, the meanings of these terms shifts with them. So we do not have an equivalence to the rules of inference.
Okay, so let's say that "energy" or "bound-state" or "force-carrier" apply only to a subset of universes. In that case, plasma is still portable across that subset of universes. BTW, physicists do discuss alternate universes with these concepts, so we're not talking about altering definitions in any meaningful way. There are strict definitions for energy, bound-state and so on that are portable across universes.
Furthermore, I could raise your own objection for universes, saying that inference is only portable within the subset of universes that have laws of non-contradiction. The fact that the subset of universes supporting inference is broader than the subset supporting plasma does not break the analogy.
Minds do not operate according to the rules of inference all the time. Let's not pretend that naturalism says that rational thinking is how all organic matter must act all the time in light of particular physical laws. Naturalism says that combining certain physical laws in certain physical configurations, you will get and inference-making system. If you change the physical laws, then you may be able to compensate (and maintain inference-making ability) by altering the configuration. This is not a problem for naturalism because inference itself is defined to rely only on a system's logical prerequisites, not on the particular laws of that system.
Let's say we assume moral realism. If there are objective moral laws, it is difficult to see how these laws could affect purely physical systems.
Or anything at all, for that matter.
But not all naturalists find it easy to dismiss all morality as subjective preference.
Are you really going to argue that an intuition (that there is moral reality) ought to trump our reasoning on the topic? If intuition should trump reason, why bother with philosophy at all?
DL
>>If I define a bunny as that small, furry, mammal with long ears, what has been explained?<<
You have explained what you mean by "bunny," what others mean by "bunny," perhaps even what you have learned a "bunny" to be, depending on the context. Take the statement, "Let me explain what a bunny is . . ." Such a statement hardly seems absurd or uninformative.
If your definitions of "naturalism" and "explanation" represent nothing more than your own presuppositions, they hardly have any value in argument. I can define the world as that collection of objects that depends upon God for its existence. You can define the world as that collection of objects that exists as a brute fact. But if these are definitions in the sense of presuppositions, then the discussion can go no farther. Obviously.
>>You cannot turn around and declare that pre-defined independence to be a discovery about the world.<<
They cannot be arbitrarily pre-defined, can they? They are objective, because there is only one way they can conceivably be. True, the operation of some processes of thought according to the rules of inference cannot be discovered the way humans discovered that the planets orbit according to the LUG. But the rules of inference can be discovered in the sense that someone can proceed from less knowledge of them to greater knowledge of them, wouldn't you say? Wouldn't you say as well that the rules of inference are genuinely explanatory of some thought processes? If so, then the world is the kind of place where some bodies of rules that are genuinely explanatory may become the objects of progressively greater knowledge without being inferred from the objects of sensory observation--at least, not inferred from objects of observation in the same way that we infer the laws of physics.
>>Okay, so let's say that "energy" or "bound-state" or "force-carrier" apply only to a subset of universes. In that case, plasma is still portable across that subset of universes. BTW, physicists do discuss alternate universes with these concepts, so we're not talking about altering definitions in any meaningful way. There are strict definitions for energy, bound-state and so on that are portable across universes.<<
>>Furthermore, I could raise your own objection for universes, saying that inference is only portable within the subset of universes that have laws of non-contradiction. The fact that the subset of universes supporting inference is broader than the subset supporting plasma does not break the analogy.<<
I should have been careful to say that there are conceivable shifts in the fundamental laws of physics that would require revisions in the rules describing the physics of plasmas. In any case, a universe to which non-contradiction does not apply is not a "possible universe" at all. To refer to a "subset of universes" that includes universes that are not logically possible does not make for a convincing argument.
No matter how "strict" the definitions of rules of plasma physics may be, they take for granted that the rules describing plasmas obtain because certain underlying laws of physics obtain. There may be more than one body of fundamental laws that will fill the bill, but not just any conceivable body of laws will do so. Further, the contituents to which rules of plasma processes apply will always be subject to description in terms of those fundamental physical laws. By contrast, the constituents of the reasoning process, such as premises, are not subject to description in terms of fundamental physical laws; we cannot even conceive of how premises could be subject to such description.
>>Are you really going to argue that an intuition (that there is moral reality) ought to trump our reasoning on the topic?<<
The principle of non-contradition is an intuition in the sense that it can neither be confirmed from observation nor proved by argument, yet we find ourselves compelled to asssent to it.
That the universe is intelligible is an intuition if Hume was correct in outlining the Problem of Induction. But again, it is an intuition that just about all of us (scientists, in particular) find impossible to reject.
There is a difference between a private intuition and one which is universally accepted, or thoughtfully accepted by a great many otherwise rational people. The intuition can still be argued, but a distinction is in order.
Relatively few people (including scientists) claim to reject moral qualities altogether and even fewer consistently behave as if they do. There is no question of the moral sense "trumping" reason if it is, in fact, reasonable to regard certain human actions as morally right and others as morally wrong.
Darek,
Take the statement, "Let me explain what a bunny is . . ." Such a statement hardly seems absurd or uninformative.
Yes, but what is being explained is our usage of the term bunny, not anything about bunnies themselves.
If your definitions of "naturalism" and "explanation" represent nothing more than your own presuppositions, they hardly have any value in argument.
They are not mere presuppositions. They are formalizations of our intuitions that
1) there is a distinction between the explanatory and the non-explanatory,
2) that restatement is not explanation,
3) that some things have no candidate explanations, let alone probable explanations.
Now, if we like, we can define your kind of explanation (whatever that may be) as a D-explanation, and mine as an L-explanation. Philosophical investigation will show us the consequences of accepting either kind of explanation. For example, if you give up (2), you also give up (3), since every phenomenon has the trivial explanation that consists of restating the phenomenon.
Suppose we are trying to explain X, and we hypothesize that there is a physical theory of X that predicts X. However, since we do not know anything about this theory apart from the fact that it predicts X, we can make no other predictions nor introduce any other constraints on observations consistent with the theory. This mere reference to a theory we do not yet have - does it have any power to explain X?
If theories we presently lack have explanatory power, then I can explain any phenomenon merely by referring to a Theory of Everything, notwithstanding the fact that I don't know anything about the details of this theory.
On the other hand, if you reject the power of reference to unknown explanations to explain, then you agree with me that prediction is a requirement for every candidate explanation.
Whether or not you agree with me is a question of value. I don't think it can be proven either way. What is not a question of value are the consequences of each definition.
I have lots more to say about our core dispute, but, if it is okay with you, I would like to start a thread on my own blog where we can discuss it. This one is getting challenging to read on my mobile phone!
DL
It seems that we have strayed quite far from the original subject in this part of the discussion. Without being tendentious, let me comment briefly on the following:
>>Philosophical investigation will show us the consequences of accepting either kind of explanation.<<
>>However, since we do not know anything about this theory apart from the fact that it predicts X, we can make no other predictions nor introduce any other constraints on observations consistent with the theory.<<
To me, the first statement of yours above is a kind of prediction, namely, that if I think hard and logically I will see the reasonableness of your formulation. However, it is not a prediction of sensory observations such as you seem to be referring to in the second statement. It is not the kind of explanation that predicts, for example, that if gravitational theory is correct I will see that canon balls of different weights fall with the same velocity. There is a prediction of a kind of observation or a "constraint on observation" in either case, but in the first instance the observation to be made is introspective while in the second case it is sensory. I believe that explanatory value can entail predictive efficacy if both these kinds of predictions are accomodated, whereas you seem to be saying that explanation must be entailed by the second kind of prediction. Excuse me if I read you wrongly on this. I will get over to your blog when I can.
Thanks Dr. Logic and Darek B for a great discussion! I'm learning so much even as a bystander. I'm wondering how you guys get the time to do this so well. Darek, do you have a blog or site somewhere on the net? I'd love to read more of your thoughts in general. Your ideas are interesting and have been very helpful. Thanks, fellas!
PB
Thx for your comments. I don't think any of us have as much time for this stuff as we would like. I don't have a blog, but if you email me I can give you some urls. My address is my first name at flexiss.net.
Post a Comment
Subscribe to Post Comments [Atom]
<< Home