A little while back, my friend UnkleE suggested that I should consider some questions that he believes are problematic for atheism as a worldview. He listed 5 questions, and I want to take them one at a time, so they can each get the focus they deserve.
That said, my initial responses to each of these questions may not be very long. Instead, I’d like to use each of these posts as a launchpad for discussion. I know these are issues that UnkleE (and probably many of you) have thought about at length, and I’d like to consider those arguments as fully as possible without subjecting everyone to my own rambling preamble. So, here’s question 1:
Do we have free will? If so, how? If not how can any choice be based on evidence rather than brain processes?
I don’t know.
I’m aware that a number of physicists and other scientists sometimes argue that free will is an illusion. That was shocking to me when I first heard it, but I now realize what they’re saying.
Imagine you could go back in time to a point where a decision was made on something seemingly insignificant. In 1959, Buddy Holly, Ritchie Valens, and the Big Bopper died in a plane crash. It’s said that another musician, Tommy Allsup, was going to be on the plane instead of Ritchie Valens, but they flipped a coin for it, and Valens “won.” If you could go back in time and witness that coin flip without interfering, would anything happen differently? Whoever came up with the idea of flipping for it thought of that for very specific reasons that would still be the same if it happened over again. Allsup flipped the coin at a specific level of force, and it flipped through specific atmospheric conditions. Those things would still be the same if you were watching it happen. Valens called “heads,” which he did for specific reasons, even if they were subconscious. In other words, every single thing that happened, even though they were seemingly random, happened in particular ways for particular reasons. If you could replay it over, there’s no reason to think anything would play out differently.
And every decision you’ve ever made, you made for specific reasons, even if the decision was close. If you went back in time and made the decision over again, but only knew the same things you knew at that moment, could you have made any other decision?
There’s no real way to test this, but the thought experiment leads many to conclude that true “free will” is not really possible.
I’m not sure how I feel about that. I do think that if you could replay decisions, it’s unlikely they would ever change. But that’s not really what I think of when I think of free will. Just because I made all my decisions for specific reasons and was “powerless,” in a way, to do anything different, that doesn’t mean that I had no control over the decisions. Thought processes were still firing in my brain as I calculated a number of factors, considered past experiences, estimated probabilities, and tried to predict possible outcomes. I might always come to the same conclusion in the same circumstances, but my mind is still very active in the process.
[H]ow can any choice be based on evidence rather than brain processes?
I think any choice — any good choice — should be using both. Brain processes deal with information, and that’s all that evidence is, so I see them as being very closely related.
Ultimately, I don’t see how this question causes a problem for atheism. I may have more to say about it in the comment thread, but I’ll need to see the case against atheism filled out a bit more before I can really weigh in on it.
Finding Truth 
UnkleE,
I do think we live in a physical world, and I do think that we are able to choose between options, but I do not think it’s libertarian freewill (at least as the definitions I’ve seen). And I do not see how a god or a hereafter would eliminate the physical constraints or influences that we have now.
I also find the whole discussion very close to being moot and futile. If we really don’t have any choices at all, and if we’re all essentially pre-programmed by each previous event in our lives, then what happens, is what happens, is what will, and only could happen – so moot.
If we do have a choice, then it’s still surrounded by our physical would, and all the prior experiences and knowledge, as well as our own ignorance, wants, needs, moods, etc, whether is or is no god. Having a god wouldn’t allow us to make decisions outside of those influences and constraints – so, to me, again, moot.
I’m reluctant to take abstract ideas and place them into neatly defined, small boxes. The exercise limits the words, but not necessarily the possibilities. We act as if the box itself is the idea, but it’s only the starting point in which we’ve begun to really contemplate the idea. In other words, our terms and the way we define them may fall too short of reality, making us blind.
And even if I’m just huge moron who just can’t seem to grasp this simple and obvious problem, surely this isn’t THE issue you hang your hat on, the one BIG issue you have against atheism… So even if I just fail to see the problem here, I’d guess that you have other issues that you feel like are problematic for atheism – I’d like to see those.
LikeLiked by 1 person
Dave: More thoughts on the “reality” of randomness. Dice-throwing “looks” random because if you measure the frequency distribution where the dice land (1/36 of the times they land “2”, 1/18 of the times they land “3” etc.) it looks just as you would expect if the landing of the dice were a completely random process.
But it’s more than “external appearances” from the “looks” of the frequency distribution, because, additionally, there’s no way to predict where the dice will land, other than using the expected probability distribution assuming randomness. In other words, you can’t improve a prediction beyond the frequency distribution. It’s the unpredictability that allows you to say that the process is truly random.
For example, sometimes it’s useful for a computer to give you a “random number” (in gaming applications, or scientific simulations, or when adding “noise” to a learning system, etc.). But these numbers are not actually random, and they’re known as “pseudo-random” number generators. They basically come from complicated equations that produce sequences of numbers that “appear” to be random because (like the dice) their frequency distribution “looks” random. However, if you know the equation, you can predict the next number in the sequence, which means it’s fully deterministic (causal) and not random.
Even if you don’t know the equation of the random number generator, you can still find it or a close approximation to it. For example, Artificial Neural Networks can “learn” the equation just by looking at examples of the series of numbers, and can approximate the underlying equation (without ever “seeing” the equation itself), just by looking at example after example of the series of numbers. Eventually, they will “learn” the pattern and be able to predict the next number with a high degree of accuracy.
Here’s an anecdote. Some years ago a smart ass won a couple of Keno jackpots in a row in Vegas. The odds against winning a single jackpot are astronomical, let alone two in a row. Keno is a table of 20 numbers from 1 to 15 and the more of them you guess correctly the higher your winnings. So this smart alec sat in a Keno lounge for a few days and fed the sequence of Keno numbers to his Machine Learning algo in his laptop. The algo eventually “learned” the equation that generates the pseudo-random numbers by the Keno computers, so he was able to guess all of them and win the jackpot (well, his Machine Learning algo did). But the idiot got greedy and did it twice in a row and got caught, and lost all his winnings. 🙂
The point is that if there’s no way to predict a process, then it is random, by definition. More sophisticated computer chips have truly random number generators by using the thermal excitations of silicon in the chip, which are truly random processes (instead of using some pseudo-random number generator based on a complicated equation). There are many random processes in nature that are truly random in that they are entirely unpredictable. Regardless of how precisely you can measure them, you can’t predict them, and that’s what we call random.
Notice that unkleE is making yet another whopper when he said this to you:
Nope. This is a false dichotomy. It’s not an “either or” it is a combination of cause-and-effect physical processes with unpredictable randomness. That combination can run the gamut from causally determined to intermediate to random depending on the situation and it is not “binary” at all, it’s a continuum. Most things in the real world are like that.
As you pointed out earlier, if it’s not causally predictable, it’s random. However, something can be intermediate and be somewhat predictable and have some randomness to it (it’s called probabilistically or stochastically predictable). This means that you can’t predict exactly the next outcome, but you can make a prediction that’s better than what you would expect from a strictly random probability distribution. There are many examples of this (the weather, radioactive decay, etc.).
No matter where in the continuum between causal and random a particular instance of decision making lies, there is no room for LFW, because it is not supposed to be either causal or random or any combination of the two. It seems that the concept of LFW is simply incoherent.
LikeLike
@William
Nailed it,William. The core issue, of course is that Unklee, and maybe many other theists, cannot grasp why we are unable to see, or as far as they are concerned, refuse to accept </em , their belief in the view that design must equal designer and all that follows from this and that the designer must therefore, be Yahweh/Jesus of Nazareth.
But we have no problems with our position, seeing no more reason to accept a fictional deity called Yahweh any more than unklee would accept Quetzalcoatl or Hanaman or Thoth.
We reject the supernatural and if theists believe there is a case to be made, then they must begin at the source of their belief – the bible.
They must have the integrity to demonstrate the veracity of its claims. To show exactly how Yahweh is Jesus is Yahweh and answer every charge against the text being spurious.
But they will not do this, simply because they cannot do this, so it is far easier to begin their defense Arse-Backwards and build a case from something the atheist can no more disprove – or demonstrate – than the theist can, which leaves them the slender opening to insert the ”goddidit” line once again.
It isn’t lying, but it isn’t honest either.
It is in fact, almost an attempt at a gotcha tactic.
And you are correct, that for us atheists the whole issue is largely moot.
But the topic does illuminate just how much someone like unklee is prepared to bend and twist normal reasoning to arrive at the position that will allow him to truly believe Jesus of Nazareth is the Creator of the universe and thus his arguments here are true.
My gut tells me that even he has difficulty fully accepting this and is simply squeezing his bottom and pushing the Concord Fallacy.
Ark.
I think it’s time Nate dropped this and moved on to something more tangible, wouldn’t you agree?
LikeLiked by 1 person
Hi Dave,
Sorry to be so slow in replying, but I have been out every night.
”So if our brains are determined, how come we think so differently?”
“Because our brains don’t develop in a vacuum being fed the exact same pieces of information.”
This is the obvious response, but I wonder if it stands up to examination?
The argument is that although the brain works by physically determined cause-effect processes, they will correlate via natural selection with ground-consequence inference.
Natural selection works by favouring those genetic inheritances that more often survive. So if we start with a population of zebras, their maximum running speed will be a bell curve from faster to slower. The slower ones will be easier prey and so those genes will reduce in the population. Perhaps the faster genes will reduce too, because they tend to expend extra energy to achieve that speed. Over time, the maximum running speed will increase, and the bell curve will get narrower. But other factors, perhaps eye colour or length of tail, may not be so critical, and so natural selection may not narrow the bell curve so much on those characteristics.
And it is important to note that the zebras have no real choice in this. They are all the result of the hereditary given to them by natural selection, and they all end up with stripes, not spots, and they all end up being able to run fast.
So, with humans, the ability to do simple reason would be advantageous to survival, so we expect natural selection to breed out those whose determined cause-effect brain processes don’t lead to logical outcomes, and the original bell curve of cognitive ability would narrow and the average ability would rise. That’s the theory.
And for simpler cognitive processes, that seems to have worked. Most people can learn a language, do simple arithmetic and know to run if they hear a lion close by. But when we get to higher cognitive processes, we are all vastly different. There is a wide bell curve of IQ. We think very differently and disagree on many important matters. When it comes to belief in God, which is supposed to have arisen through agent detection, we disagree profoundly, we can’t even agree on what is evidence, how much there is and how we can draw inferences from it. The two sides (as evidenced on this blog) can hardly understand each other. We clearly think very differently.
It certainly looks like something other than natural selection has been at play, as you agree with your answer above.
But only natural selection can produce the correlation between the cause-effect brain processes that are all there is if determinism is true, and the ability to do inference and ground-consequence logic that we all admit human have, in some measure. So it looks as if what we observe hasn’t been produced that way, and we still lack any realistic explanation of how our cause-effect brains can do higher inference and thinking.
Worse for determinism, if your explanation is correct after all, then the same natural selection that produced your brain produced mine, despite the differences in our thinking, and there is no way to judge which has the better correlation between cause-effect processes that actually occur and the ground-consequence outcomes we hope result. You will think one way and I’ll think another.
And yet we are discussing as if we can somehow make that judgment.
”It appears you are appealing to mystery, that the origin of the choice “cannot be explained”, rather than providing an example of a choice that is made for a non-random, non-determined reason.”
Not at all. I am saying we all know and experience what it is to make a choice. We have many inputs, some are causally determinative or impossible (I can’t fly and I will fall if I jump off a cliff) but others of them are matters of judgment or taste. There are reasons for and against each one of these, and we may each assess them differently. For example, if choosing a meal at a restaurant, I may prefer meal A, but I wouldn’t mind trying B for a change, and while I like C I remember I got indigestion last time. The choice certainly isn’t random (there are good reasons for and against each choice), but it isn’t determined either. I will choose one way in the end, but I could easily have chosen a different way.
So the experience is universal, and while you as a determinist may disagree, my explanation makes sense and presents something other than determined or random.
If belief in God was the result of natural selection and agent recognition, you’d expect everyone who survived to have had some sort of numinous belief. And it appears that for millennia that may have been the case. But as the human brain and culture have attained greater cognitive powers and thinking, we can profoundly disagree and many of us (e.g. you) have thrown off the thinking that natural selection gave us. Which again suggests that something more than natural selection and determinism is going on when we get to higher cognitive processes.
”Can you think of a choice (judgment) that comes from a non-random, non-determined source?”
So yes, I think my choice of a meal, and your choice of atheism, are both examples. And millions of other things too.
LikeLike
Hi Travis, hope you had another good few days out in the wild.
”Let’s forgot about the major point for now because if we can’t get to mutual comprehension on the simplest of examples then discussions about scaling up to more complex situations aren’t going to go anywhere.”
Yeah that’s good.
”I don’t follow your response. As best I can interpret, it sounds like what I noted in my previous comment – that you object to the sufficiency of this model on the grounds that associations can be formed regardless of whether they properly reflect real-world associations. Is that the issue?”
Let’s start with your original example (I have added numbers for clarity):
”All men are mortal. Socrates is a man. Therefore Socrates is mortal.” (1) If we assume that our model comes to this syllogism with a relatively blank slate, only previously containing the concepts of men and mortality, then the model would not presuppose that the new entity “Socrates” has any established association with mankind or mortality. The conclusion will be a completely novel prediction. (2) Upon receiving the semantic content of P1 (All men are mortal) the network creates a new association between the existing concepts of men and mortality. Upon receiving the semantic content of P2 (Socrates is a man) the model has created a new concept for Socrates and a corresponding association to the concept of men. (3) The net result (no pun intended) is a new association between Socrates and mortality by way of the associations of the intermediate concepts as they were introduced into the model. (4) If we are to then prime this model with the semantic content of “Therefore Socrates is…” the predictive paradigm would build upon the fresh association to predict the mortality of Socrates via the association.”
(1) We are starting with a blank slate.
(2) The network creates new associations, but because of (1), it has no idea at this stage whether these associations are true or not.
(3) So yes, there is a new association between Socrates and immortality, but if a different input had been received, the opposite association would have been formed, because at this stage the brain is a blank slate. We will know that “Socrates in immortal” is a faulty conclusion, but this blank slate brain will not.
(4) So yes, this will happen, as it equally would happen (but with opposite conclusion) if the input was opposite.
Now how will the blank slate brain be able to decide between good and bad inferences? If the person sees Socrates dead, then the brain would verify the correct answer. But for higher cognition, and even for lower cognition (I haven’t met Socrates) we usually don’t get that opportunity.
So to attain the correlation between cause-effect and ground-consequence, we need to have used that correlation, which of course is circular.
That is what I was suggesting.
LikeLike
Dave: More confusion by unkleE; same tired old statements that have been challenged, but the challenges were never addressed by unkleE. The man simply keeps ignoring devastating criticisms and keeps repeating himself, treading the same old stuff. So here we go again.
This is a one-sentence conflation of ontology with epistemology, the unkleE Two-Step Shuffle.
Nope. Natural Selection does not favor survival, it favors reproduction. Now, survival is very often tied to successful reproduction, but not always. A peacock’s tail is disadvantageous from a survival perspective, but it evolved because of sexual selection (preferences of pea hens). Sense of humor, risk-taking, resourcefulness, craftiness and intelligence could well have been sexually selected in humans, besides also being demonstrably advantageous in survival and reproduction for reasons other than sexual selection.
This is overly simplistic, as there are likely many tradeoffs involved with speed and reproduction, (like energy conservation, bulkiness of certain muscles needed for speed but interfering with other anatomical functions etc.), and that’s why speed doesn’t continue to increase ad infinitum. But the bell curve does NOT have to become narrow. In fact, a broader bell curve has advantages because it confers the population with flexibility and plasticity (look up “evolvability” and “phenotypic plasticity.” Genetic diversity is in fact crucial for long-term species survival. The cheetah, for example, is in danger of extinction, not because of the size of its population, but because it is highly inbred due to having undergone a relatively recent population and genetic bottleneck. So unkleE is just pulling the “narrowing of the bell curve” out of thin air. Genetic diversity is generally advantageous to the long-term survival of populations, because it allows them to trade off different traits as the need arises over geologic times.
No, that is a cartoon of the theory. First of all, there is no “narrowing of the bell curve.” Second, notice unkleE’s fetish with only “simple reason” being advantageous to “survival” (or, rather, “reproduction”). But why must it only be “simple” reason? Why draw this artificial boundary where only the simplest of cognitive processes can be advantageous to reproduction and not more complex ones, particularly in a social and highly interdependent species like humans? unkleE doesn’t explain this, he assumes it.
The answer is simple. There’s no need to draw this artificial boundary between “simple” cognitive processes and more complex ones being selected for by NS. It seems that unkleE does this, because it gets him to the result he so desperately wants: “Higher cognition cannot be selected for via natural means, and instead requires miracles and magic, and therefore, “god-of-the-gaps-diddit.”
Hmmmm, let’s see. NS worked for “simpler” cognitive processes, but no explanation whatsoever as to why it couldn’t have worked for more complex cognitive processes. This is an artificial boundary that unkleE is drawing out of thin air without any explanation. This is the same tired old canard.
Again: unkleE needs to demonstrate that NS could not have selected the more complex cognition because either (1) higher cognition has no reproductive advantages whatsoever for a highly social and interdependent species, or (2) it could have never been selected by NS even if it were advantageous for reproduction. unkleE has not answered this challenge, but continues to repeat the same canard.
Doing “simple arithmetic” and “learning a language” are NOT simple little cognitive tasks. They are already highly complex and require a tremendous level of complex cognition. Now unkleE wants to lower the bar of cognition and lump these with “running when you see a lion” so he can sneak in his “simple” label and dismiss the fact that more complex types of cognition can, in fact, be selected for by NS.
We are all “vastly different…” First of all, we are NOT “vastly different.” We are very much alike in many ways. We agree on many things, and we suffer from pretty much the same cognitive biases across the board, exactly as we would expect from naturally-selected cognition. We do have some subtle differences and disagree on many things, but so what? This is the unkleE Theorem again, which has been debunked over and over. There is no problem with “disagreement,” it is expected under Naturalism and under Determinism. It is less expected under Theism where we supposedly have an “external guide.”
On Theism, why would we disagree so strongly on, for example, moral questions that are supposed to be “grounded” on some invisible daddy figure in the sky? On Theism, why would we disagree that there’s an invisible daddy figure in the sky when he loves us so much and wants nothing more than for us to know him, and to have a “personal relationship” with us (violin music fades in and out)? “Disagreement” is perfectly natural and expected on Naturalism or Determinism, but it is not expected on Christian Theism. Therefore “disagreement” is a problem for Theism and not for Naturalism or Determinism.
That is simply false. We have a perfectly reasonable and realistic explanation of how this might have come about. It’s called Evolution by Natural Selection, and there is a vast literature on the subject of the evolution of the brain and intelligence in human beings, no “magic” required. This recalcitrant incredulity and denial is breathtakingly ignorant, particularly with the little effort that unkleE has clearly put into understanding this, and his constant cartoonish depictions of evolution. Moreover, it is singularly rich, since his god-of-the-gaps explanation of “magic,” is not an explanation at all, but a combination of a Fallacy of Argument from Ignorance, and substituting a “mystery” for another “mystery.”
Again with the “disagreement” canard. So what? Disagreement is expected under Determinism and unkleE’s Theorem is unsound, no matter how many times unkleE repeats it. The “ground consequence outcomes that we hope result” don’t have to result always, and often don’t result. To expect this is either a naive reification of “logic,” or wishful thinking.
So what? Our internal subjective experiences are NOT reliable pathways to objective truths. This is the same “feelings, nothing more than feelings” canard recycled yet again. unkleE has been challenged REPEATEDLY to explain either that (1) “feelings,” “impressions,” “introspections,” and now “experiences,” are 100% reliable 100% of the time, or (2) that they are 100% reliable in the specific case of LFW. He has not met this challenge, but continues to repeat the same tired old canard that “they just are reliable, because unkleE says so.”
The choice may have some “random” component along with some “good reasons for and against.” But how does unkleE know that the choice was not some combination of randomness with something determined? ,/strong> How is unkleE able to rule out either or both or a combination of these? How does unkleE know that he could have done otherwise if we had a time machine and could “replay the tape”? Has unkleE run this experiment all by his lonesome and know something the scientific community doesn’t? Or does he just have a “feeling” about this? This is precisely what we are debating here and unkleE simply re-states his pet conclusion and declares victory! What does unkleE propose drove his choice if it wasn’t some combination of “random” and/or “determined”? “Magic”? What does that look like? How does that “magic” interact with the physical world to make him choose his decision and express it via his brain and nervous system? Does unkleE even understand that he is begging the question here?
Unfortunately, unkleE has offered NO explanation, but a continual recycling and repetition of tired old canards that have been challenged and debunked. This is what unkleE has offered so far:
(1) An equivocation between ontology and epistemology where he hops back and forth between the two. This is the unkleE “Two-Step Shuffle.” When the first one is challenged, unkleE half acknowledges the validity of the challenge, but hops to recalcitrant incredulity of the second one. When the second one is challenged he half acknowledges the validity of the challenge, but hops back to recalcitrant incredulity of the first one which he had half-acknowledged before. Repeat over and over.
(2) The “feelings, nothing more than feelings” canard. unkleE has been challenged to demonstrate that “feelings,” “introspections,” “experiences,” and the like, are reliable pathways to truth. He has ignored this challenge, and continues to repeat the canard. All he does is change the word from “feelings” to “introspection,” etc., and the latest iteration of the word is “experiences.” I predict that pretty soon he’ll run out of synonyms and start at the beginning with “feelings” again. Or maybe he’ll go back to the “cognitive dissonance” psycho-diagnosis.
(3) The unkleE Theorem, or “we all disagree” canard. That this is not a problem for Determinism has been explained ad nauseam. Plus, unkleE has been challenged repeatedly to explain how examples from Artificial Intelligence, being deterministic, can “disagree” among each other, but he has ignored this challenge and continued to repeat the tired old canard. Disagreement is expected on Determinism and Naturalism and NOT expected on Christian Theism. To quote that great philosopher Rickie Ricardo, it seems unkleE has a lot of ‘splainin’ to do here.
(4) unkleE has re-stated his pet conclusion that “it’s neither random nor determined” over and over without proof, in a perpetual and circular “begging the question.” He doesn’t seem to notice that he can’t just state his conclusion and declare victory.
Not at all. Belief in gods may well be one of our many naturally evolved cognitive biases. Type I errors, including hyperactive agency detection, are known to be favored by evolution when taking into account a cost-benefit analysis; look up “evolution of superstition.” So, beliefs in gods are expected on Naturalism and on Determinism. However beliefs in hundreds of thousands of different gods throughout human history is NOT expected on Christian Theism. More ‘splainin’ for unkleE to do here. And yet again, “disagreement” is not a problem for Naturalism or Determinism; it is a problem for Christian Theism.
Yes, unkleE can think of many things, if his thought process is wrong, but I’d say: Prove it. This is what’s under discussion here, and unkleE simply asserts it while incurring a ton of fallacies in his non-explanation. Again, how does unkleE know this???
LikeLike
Dave: So sorry, I think I messed up a few of my “boldface” tags, so there’s a lot of bold in my comments–didn’t mean to be confusing with overusing boldface.
LikeLike
Travis: You asked me to chime in, so here are some thoughts/clarifications. When you say that “The conclusion will be a completely novel prediction,” referring to the “Socrates is mortal” syllogism, actually, in deductive logic conclusions are not “novel” in the sense that they don’t represent new information. Conclusions in syllogistic logic are already contained in the premises. The process of syllogistic logic can only give you “valid” arguments in that, if done properly, the conclusion follows inescapably from the premises. However, this says nothing about whether the conclusion is “true” or “false.” If the premises are false, or are not demonstrated to be true, then even if the argument is valid, it would be “unsound” and the conclusion cannot be trusted to be “true” from the argument alone. Logic is a closed system that generates no new information. If we want new information, we need to, observe the world to come up with new premises that are substantiated by observation.
Also, I don’t think brains are “blank slates.” We are born with a great deal of pre-wired information and a bunch of pre-dispositions. For example, newborns are very adept at recognizing faces and even facial expressions, and are pre-wired for learning language, a sense of “fairness,” and a host of other behaviors. The pre-wiring can be explained through Natural Selection, and subsequent learning through cultural evolution. There’s no need to invoke “magic” much less “god-of-the-gaps-diddit” in any of this.
Moreover, logic, mathematics, etc. don’t come easily to us one bit (I can attest, having both learned and taught it). Our brains are, at best, logic “emulators,” riddled with fallacious thinking, heuristics, rules of thumb, and short-cuts, which work well for many of our day-to-day tasks, but which can leave us with huge blind spots when it comes to pure logic or math. That’s why it takes years of training to become adept at logic and math with any kind of sophistication.
If you’ve ever looked at a mathematical or logical proof beyond a pet example, it uses many “baby steps” to prove “large steps.” A theorem is a large leap of logic which is substantiated with many intermediate baby steps. This is because we don’t immediately see the large leap, but we can follow the smaller intermediate baby steps. If we were really good at logic/math, we would have no problem seeing the large leaps without needing the baby steps in the middle (you can develop a “knack” for this after many years, but not many people can do this, and they still need to double-check their hunches with baby steps). A theorem always follows from the axioms and its conclusion is nothing new. The reason a theorem is valuable is because we are so lousy at discovering all of the possible implications of a set of axioms immediately, so we need to state the large leaps explicitly. If we were “omniscient reasoners,” all we would need is a set of axioms or premises, and all possible implications, theorems, etc. would be immediately known to us. But we are not very good at logic and math, and our brains have to be coaxed into doing these things kicking and screaming and with a lot of training.
We’re adept at “filling in the blanks when we have faulty and/or incomplete information, and we’re good at reaching quick-and-dirty decisions which are right most of the time (those times that matter to our survival and reproduction), but completely wrong other times, and we’re not so good at deep thinking, unless we are trained to do so, and we observe and double-check with others around us and across generations, and observe the real world to help us double-check our cognition. This is precisely as you would expect from a brain and cognitive system that has been cobbled together by Natural Selection.
As to “cause-effect” physical processes being able to produce “ground-consequence” logic of the level of the Socrates syllogism, yes, of course they can, a million times over. There are AI systems that, given a set of axioms and premises, can generate deductions and theorems that are hugely more complex than the Socrates syllogism. Look up “Automated Theorem Proving.” These systems are routinely used to aid human mathematicians with “discovering” new theorems, etc. Therefore, ontologically, yes, physical processes can demonstrably give rise to very sophisticated logic. The reason why I use AI examples is because no one can deny that they are “cause-effect” deterministic physical systems, which addresses the “ontology” question of “is it possible for physical systems to encode logic.” The answer is a resounding “yes.”
Faced with this undeniable fact, someone who frequents this blog, will immediately shift his “recalcitrant incredulity” from the ontological question to the epistemological question like this: “Yes, but that’s because they were “programmed” by a human,” who of course, got his cognition from a “magical” source in turn.” But the question of epistemology–or origins–is separate from whether “cause-effect” physical systems can produce “ground-consequence” logic (they can). The epistemological (origins) question can also be satisfactorily addressed through Natural Selection. AT this point the person in question will invariably offer “cartoonish” depictions of evolution, along with misunderstandings and flat out falsehoods, and assert, without proof, that only “simple” cognition can be selected for, but not more complex cognition, because “cause-effect” could never yield “ground-consequence,” and we’re back to the ontological question. Repeat over and over.
Speaking of that someone, unkleE has now contradicted himself. Remember how very fond unkleE was of pure “introspection,” and how he said that we should trust “introspection” to guide us to the truth (of whether we have LFW), yet now he has admitted that a syllogism can be wrong without observation of the real world. In other words, he’s now saying that we have to observe Socrates to know that he’s in fact mortal and not just go by the “introspection” coming from the pure syllogism, because the conclusion could be wrong if the premise (observation) is wrong. Well, I’m glad that unkleE is coming around to that reality, finally. I suspect, however, that he’ll go back to “introspection is king” when it suits him.
We don’t start out with a “blank slate brain,” but regardless, the way you decide between good and bad inferences is through, ahem, observation, and cross-checking and learning from others. Human children have one of the longest “apprenticeships” in the animal kingdom. Their brains spend a lot of time observing the world and cross-checking against more experienced brains.
We don’t have to have met Socrates (or Jesus), we can use inductive inference to know that every single person we’ve ever known of in history has died and it is extremely likely that all persons alive today will die after a certain age. We don’t see anyone alive today who was around during or before the American Civil War, for example. The human lifespan world record is, I believe, 122 years, unless you care to believe the fable that Methuselah really lived to the ripe old age of 969, which stretches credulity beyond any reasonableness. But even Methuselah, if he existed, was said to have died. We also have never seen a corpse come back to life, and have never had sufficient evidence to believe that this has ever happened in history (despite myths and tall tales to the contrary, but the evidence just isn’t sufficient).
This simply does not follow. unkleE’s entire comment is a morass of confusion between deductive and inductive logic, along with the ontology vs. epistemology question of whether logic can in fact be encoded in physical systems, and how that can arise.
The point is that “the input is NOT opposite,” so there’s no circularity here.
If the inputs were incorrect most of the time, then yes, we would reach unsound conclusions, but if the inputs were correct most of the time, we would reach mostly sound conclusions. What we observe is that we have a mixture of correct and incorrect conclusions based on correct or incorrect observations. We see agreement with the real world (and each other) sometimes, and disagreement with the real world (and each other) sometimes. The fact that in some respects we see more agreement than disagreement with the real world is testament to how our cognition reflects the correct “inputs,” albeit imperfectly.
The epistemic “correlation” between “cause-effect” physical processes and “ground-consequence” logic can produce valid conclusions from premises (in other words, it can produce the encoding of rules of deductive logic in physical systems, which is demonstrably possible). Observation of the world gives approximately “true” premises, which yield approximately “true” conclusions (inductive logic). To the extent that this has been advantageous for reproduction in our evolutionary history, this is to be expected and is in fact anticipated. There is no problem here.
LikeLike
Travis:
“I am also willing to move on to “the larger problem” but it’s pretty important that we first agree on what it is that you are granting. From my perspective, the framework being granted is that there are viable cause-and-effect models of learning networks that are generalizing prediction engines which build probabilistic associations from inputs (evidence) to produce outputs (predictions) which tend to agree with reality. Is this what you’re granting, even if only for simple cases?”
I said back on June 21, about cause-effect processes producing ground-consequence logic, and I am saying now, that “let’s assume that it can, for the moment, and see where that leads.” I am not granting that it can happen for complex cognition, though I accept it can for simple cognition, but I was, and am, willing to grant it for the sake of argument so we can explore the later “problem”. So if we want to discuss what I regard as the major difficulty, I am happy to assume for the sake of argument that all levels of cognitive inference can arise from cause-effect brain processes.
“lead me to think (again) that you are assuming logic to be a separate, transcendent entity rather than something that naturally arises in the framework do to its reliance on associations.”
No, I have never really thought about such an assumption, I don’t think I even know what that statement means. All I am saying is that if the ability to do logic naturally arises, then it is governed by natural selection, which selects via survival and reproduction ability, not logic (though logic may lead to survival). I am suggesting that it isn’t enough to assert this happens (like a just so story) but to show how it happens. And to live with the consequences of the process you have suggested (which is my main point).
“I propose that the onus is on you to explain why the model becomes unreliable as it scales to more complex issues”
I don’t feel any onus. I am simply asking for an explanation of how this process produces the outcomes we actually see..
“Perhaps ‘unreliable’ needs to be defined?”
I mean not as reliable as we actually experience. It seems as if we can generally think very reliably even at a very high level. For example, Andrew Wiles found a proof of Fermat’s Last Theorem, it took 129 pages to write out. He found he made one mistake (just one!) and had to work to resolve that difficulty, and he was able to do it! To conceive of such a proof and to be able to write it out so accurately is an enormous achievement, way beyond simple syllogisms. Scientists, mathematicians and philosophers often report their work starts from an inspired insight which then have to develop into a proof or demonstration. It is hard to see how this level of cognition, which is beyond almost all of us but nevertheless produces such precise results, could evolve by giving humans a survival/reproduction advantage.
LikeLike
It is hard to see how this level of cognition, which is beyond almost all of us but nevertheless produces such precise results, could evolve by giving humans a survival/reproduction advantage… but it isn’t impossible.
Therefore, we cannot conclude ‘Free Will did not evolve naturally’, and ‘Free Will did evolve naturally’ remains possible.
LikeLike
unkleE says: “about cause-effect processes producing ground-consequence logic… I am not granting that it can happen for complex cognition, though I accept it can for simple cognition…”
This is the same old recycled stuff, but here we go again. Exactly what does unkleE mean by “simple vs. complex cognition”? He’s given examples of “simple” as being herky-jerky reflexes like “running away from a lion.” Then he inexplicably amplified this “simple” cognition to include “language” and “arithmetic,” both of which are extremely complex cognitive abilities limited largely, if not exclusively, to humans.
unkleE needs to define precisely where the demarcation between “simple” and “complex” cognition lies, and, more importantly, he needs to define why he’s drawing an ad hoc arbitrary boundary between what can and cannot arise by natural means. He has not done this, he has only expressed incredulity and stated it as a bold unsupported assertion. His incredulity and bold assertions are irrelevant to the facts of the matter being discussed here.
Moreover, there is abundant evidence from AI that physical processes can and do reproduce logical inferences. Unless unkleE is saying that these machines have some “inspired magic” in them, the “magic” part is not necessary. So, ontologically, it is possible for physical processes to produce logic and no amount of denying this will make the evidence go away. Facts are stubborn things.
If unkleE next switches to “well, they were programmed,” then unkleE needs to separate ontology from epistemology once and for all and decide which one of the two he has a “problem” with instead of confusing them.
unkleE says: “All I am saying is that if the ability to do logic naturally arises, then it is governed by natural selection, which selects via survival and reproduction ability, not logic (though logic may lead to survival). I am suggesting that it isn’t enough to assert this happens (like a just so story) but to show how it happens. And to live with the consequences of the process you have suggested (which is my main point).”
Now this is the epistemology question. First, NS does not need to select for “logic,” or cognition, directly. Cognition and logic ability can be selected for indirectly to the extent that cognition can benefit reproduction, which it clearly can. Again, unkleE has not demonstrated that
(1) “Higher” cognition cannot benefit reproduction, or
(2) “Higher” cognition can benefit reproduction, but it is off-limits to Natural Selection.
Until unkleE demonstrates either (1) or (2), it remains eminently possible that NS can in fact select for higher cognition, recalcitrant incredulity or not. Moreover, I would point unkleE to the many articles on the evolution of the brain and intelligence in human beings. So the onus is, in fact, on unkleE to disprove all these studies. For starters, unkleE would need to explain how it is that, if the brain and cognition never evolved in humans, there is a clear increase in brain size in hominid fossils throughout human evolutionary history. These are not “just so” stories, like the fables of the Bible, for example. These are models that are rooted in facts and evidence, and that have been highly successful in explaining many observations.
As to “living with the consequences,” there are many of us who do, but this is entirely irrelevant to the truth of the matter here, and is a fallacious Argument to Consequences.
unkleE says: “Andrew Wiles found a proof of Fermat’s Last Theorem, it took 129 pages…To conceive of such a proof and to be able to write it out so accurately is an enormous achievement… Scientists, mathematicians and philosophers often report their work starts from an inspired insight which then have to develop into a proof or demonstration. It is hard to see how this level of cognition …could evolve by giving humans a survival/reproduction advantage.”
If this is “hard to see” for unkleE, it’s because he hasn’t tried (or won’t try) harder.
Clearly, unkleE doesn’t know how “theorem proving” works. Having done it myself, I can say that it’s a very messy process. Sometimes you start out thinking you can prove X starting from A. You start from A and you go “baby steps” to B, then to C, etc. and you wind up in Q instead of in X, and call it the “Q Theorem” instead of the “X Theorem.” Other times you set out to prove X and are able to do it, by cobbling up many different ideas. Sometimes you start in the middle and work towards the “beginning” and the “end.” Sometimes you start from the “end” and see if you can get to a “beginning” comprised of already-accepted axioms/propositions, and if you can’t, then you make up an “axiom” that you’re missing to get you from “axioms” to where you want to go because that’s where you wound up in the end.
That this is such an “accomplishment” only demonstrates how lousy we actually are at doing logic and higher-level math. If we were so good at this, (1) Why does it take so much training to get there? and (2) Why can’t we immediately “see” both the truth and proof of Fermat’s Last Theorem right away, instead of having to wait centuries for enough of the pieces to be put together in “baby-step” proof format for us?
Again, there’s no reason why NS couldn’t have indirectly selected for a general-purpose information and cognition processor that can produce higher level cognition, especially when you couple evolutionary biology with language and other forms of cultural evolution. Andrew Wiles, after all, is himself standing on the shoulders of “giants” who go back many centuries.
LikeLike
@AR
It warms the cockles of my heart to note that unklee has stopped responding to your comments, as it demonstrates that he has no way of refuting what you write without exposing the nonsense of his arguments.
But I am reading your replies and enjoying watching him having his arse handed to him on a plate.
LikeLiked by 1 person
It’s all pretty simple. UnkleE is a proponent of the “god of the gaps”, though he may well deny that. So he looks to what he sees as gaps in our knowledge, as a source of argument for his god.
This is not just unkleE. They all do it. That is to say, all Christian apologists do this. Most of them deny that the are arguing a “god of the gaps” — they appear to be oblivious to the fact that they are doing exactly that.
LikeLiked by 3 people
The brain is a physical thing, that exists in the natural world.
Our brains are different than a reptile’s brain or dog’s brain or a horse’s brain, and with each brain, the level of cognition has a correlation to brain type (size, construction, etc).
Doesn’t this indicate that our level of “awareness” is physical and natural, that it has a direct correlation to our brain size, construction and function – and that the supernatural element is purely conjecture?
I mean, why is it that children do not have the same level of understanding or judgement as adults? Is it because god isn’t working supernatural enough in them, instead works more supernaturally in adults? Or is it maybe due to the fact the brain just isn’t as fully developed in children? It would be fascinating to learn that god only grants freewill in proportion to a person’s natural brain ability – wow, what a wonderful and lackluster, confusing miracle.
It would be like God giving Samson strength, but only in proportion to his natural muscle mass. “Yep, God gave me strength, and all I had to do was workout 6 hours a day, 6 days a week, for the last 25 years – and now, while I’m stronger than everyone else, it’s only in direct proportion to my body mass, and my neighbor can still bench the same percentage of his own body-weight as me…” The miracle loses some of its sparkle.
But If our brains were identical to a fish’s brain, while we still acted like humans and fish still acted like fish, then the supernatural element that’s trying to be pushed here would have more weight behind it, right? Yet that’s not what we find.
But as it is, our brains are quit a bit more complex than that of a fish, or dog, or even an ape. Our brains our physical and natural and there is still a lot that is unknown about how it works exactly.
And I think we still have to take a step back and think on this: that so far, only physical, natural things have been proven and verified, so to use a supernatural position as a default guess is a very shaky position. Plus, the “supernatural” is not bound by any rules, therefore anything can literally be dreamed up as a supernatural solution – there’s lightning because Zeus is angry, or thunder because Thor is banging his hammer, there’s a god for everything and we have all the gods because Cronus vomited them up after eating them…
“in the physical world, without god, there is only cause and effect, there is no choice because we’re all essentially programmed to act and react in one specific way…” – that is just a claim, and one that ignores the actual, physical differences in brain sizes and types, and fails to explain how or why the existence of any god would or could somehow change anything about that, or alter or liberate how we select between options.
LikeLiked by 2 people
Ark: Glad to know you’re entertained. It’s only fair, as you’ve provide tons of entertainment and are one funny bloke y’self. 🙂
LikeLike
AR,
Thanks for offering your input. Here are a few clarifications that may aid you, Eric and others in understanding the approach I’m taking:
1) I fully agree that we do not start as blank slates and that large, complex problems are not naturally easy. I am only trying to work within a simplified model for the sake of discussion, in the hope that the general principles established in this simpler context can then be understood to sufficiently support extrapolation to more complex and messy real world. The model I’m offering is a non-specific entity that may plausibly execute a minimal version of the kind of learning and cognition that occurs in humans.
2) My assertion that the conclusion reached by the model would be novel was intended only to be novel from the perspective of the model. The syllogism itself may contain all the information to support the conclusion, but the model receiving that information did not already contain all the concepts and relations. The intent is to show a mechanism by which the concepts and relations in the syllogism can become realized in the model, such that the conclusion naturally arises as a result. In this way, the premises are playing the role of inputs (evidence) that would otherwise be realized through a lifetime of observation in the more complex real world situation.
I fully agree with your statement that “If the inputs were incorrect most of the time, then yes, we would reach unsound conclusions, but if the inputs were correct most of the time, we would reach mostly sound conclusions.” and I it appears to me that this is the feature which Eric finds problematic in the simple model I’ve put forth. You and I agree that this isn’t a problem and I don’t yet appreciate why Eric thinks that it should be a concern – if that truly is his position.
LikeLiked by 1 person
Eric,
We have diverged into two separate threads so I’m going to bring these back together. First let’s zoom in on the simple model and the problem you suggest:
If I understand your objection correctly, I think we actually agree: the proposed brain will reach wrong conclusions on the basis of wrong input. So let’s expand the example a bit to see why I don’t think this is a problem. Suppose that the blank slate model is fed the wrong premises of “All men are immortal” and “Socrates is a man”. At this point it generates, as you noted, the wrong conclusion of “Therefore Socrates is immortal.” Now suppose that the model is then fed the correct premise “All men are mortal” two times. This new evidence updates the associations and now the men mortal association is stronger than the men immortal association. The prediction thus updates to generate the conclusion “Therefore Socrates is mortal.” by virtue of the stronger association. At this point the model is only moderately certain about the conclusion because of the balance of evidence, but if the correct data is received more and more then the correct conclusion is strengthened. The formalization of logic takes these probabilistic associations and turns them into binary relations, but I think that the essentialism and binary relations found in logic is actually a shortcut that we have favored mostly because it reduces cognitive load (or it could be related to threshold potentials). Regardless, the net result is the same if we equate logical entailment with “the most probable output”.
So this means that the model, in general, generates predictions that build upon all received data to minimize surprise, which corresponds with the most probable outcome as dictated by that data. As I see it, this is a pretty good mechanism for reaching truth. Of course, this is also a gross oversimplification of what’s really happening, but the hope is to elucidate a general principle that supports truth-directed cognition (to the extent that the data is aligned with the truth), even if there isn’t a truth-directed teleology at the core.
Now let’s zoom out. I previously said that
Note my description of the more complex scenario says nothing about selection. The simple model still applies and is utilized for scenarios involving lots of inputs and associations. I suppose you might question why we would evolve to support the larger network of concepts and relations but recall that this is a general purpose machine. The same framework which allowed Grog to infer a spear from the prior experience with the cutting utility of pointy rocks, the action at a distance utility of long sticks and the binding utility of vines may very well be the framework involved in the solving of Fermat’s Last Theorem. The 129 pages and seven years of dedication are testaments to the limitations under which that work progressed. It required detailed step-by-step records, data from other others (past and present) and many mistakes, reviews and corrections. Only once all the data aligned to eliminate surprise was the set of concepts and associations put together by Wiles granted the esteem it holds.
Finally, I agree with William’s recent comment that “If our brains were identical to a fish’s brain, while we still acted like humans and fish still acted like fish, then the supernatural element that’s trying to be pushed here would have more weight behind it, right?” But I also admit that I do not have an easy way to identify the point at which the brain appears to be exceeding the capabilities of its cause-and-effect processes (and neither do you). All I can say is that by working from basic principles of probabilistic inferences in a neural network type model, the massive scale of the human brain, with its 86 billion neurons and each with thousands of connections, looks to be completely compatible with the cognition we have. Eventually somebody more capable than me will probably do the analysis to prove it.
LikeLiked by 1 person
Travis: OK, thank you,.I understand that you’re starting from a simple example and illustrating, step-by-step, how a simple cognition task can be arrived at, and then extrapolating from the simple example to how more complex tasks can also be arrived at in a similar manner.
What I’ve done is skip the step-by-step details that you have valiantly taken on, and instead simply pointed to extant examples of how “ground-consequence” cognition actually does happen in an entirely deterministic “cause-effect” manner in Artificial Neural Networks. These examples obviate the need to go through the step-by-step process of how this might happen, since the examples are undeniable facts, no matter how recalcitrant or incredulous the denier is. They exist, and cannot be ignored or wished away, despite the ineffectual and ax-grinding dismissals we’ve seen coming from your interlocutor. Their existence completely destroys any “ontological” objection that “cause-effect” physical processes could never give rise to “ground-consequence” cognition, whether they’re “complex” or not, and whether they might “disagree” or not.
In fact, ANNs can learn extremely complex cognitive tasks (way beyond the “toy” Socrates syllogism), and can actually discover patterns that humans had not anticipated, and can exceed humans in many cognitive tasks. ANNs are guaranteed via a mathematical theorem (Hornik, Stinchcombe & White, 1989) to be able to learn any non-random cognitive task of any complexity to an arbitrarily high degree of accuracy, if they’re given enough “neurons” and enough “examples” to work with. To the extent that their accuracy doesn’t reach 100%, they can, and do, “disagree.” There’s no contradiction here.
ANNs are flexible, general-purpose “cognition approximators” where the cognitive task is not directly programmed in, but is “learned” by the network when presented with examples. This is key, because there is no human programmer or designer expressly intervening by putting in the cognitive task using explicit “if-then” rules. The network arrives at the cognitive task by itself via “learning by example,” and can then extrapolate or generalize its learning to cases or examples that it has never encountered before.
As to the “epistemology” question of how a similar biological network can arise by natural means, notice how the capability of self-learning by example (not by direct programming or “design”) lowers the bar immensely for any kind of naturally-selected “design.” This is essentially the same as how NS has lowered the bar for abiogenesis, the complexity of the eye, etc.. For abiogenesis, all that’s required are self-replicating molecules instead of full-fledged organisms to kick-start the gradual ascent over geologic times up to the complexity of life that we see around us today. For the eye, all that’s required is a slight advantage conferred by a photo-receptive cell, then a few of those cells lumped together forming a primitive eye, then a bundle of them forming a cavity able to detect the direction where light comes from, then a cavity with a pinhole for focusing light, etc., all the way to the several different “designs” of advanced complex eyes that we see today (alongside many examples of more primitive “intermediate” eyes that we also see today).
Likewise, the initial, “unlearned” state of an ANN is simply a bunch of simple nodes (or “neurons”) cobbled up together in a random way. In other words, the simple nodes are all connected to one another with initially random “synaptic” connection strengths. When such a randomly initialized “unlearned” network is presented with an initial set of inputs, it generates an initial “garbage” or nonsensical output. Then, it is presented with what the output should have been (the correct “target” output). This slightly “nudges” the connection strengths in the direction of the correct target output. As this is repeated many, many times over with different inputs and desired target outputs, the network connection strengths (analogous to “synaptic” strengths in biological networks) will evolve and change to end up “learning” the desired cognitive task.
I’ve glossed over many details on ANNs here (e.g., I won’t go into how Evolutionary Algorithms are used to “train” ANNs, among many other examples). But suffice it to say, that it becomes quite easy to see how cognition might have evolved in biological brains through NS, similarly to how it happened with other advanced structures like complex eyes: By gradual improvements conferring gradual advantages in survival and reproduction, along with unforeseen “by-products” that may not have been directly selected for.
Neurons are thought to have evolved from specialization of nerve cells. This was followed by small bundles or “nets” of these cells like in jellyfish. Then on to slightly more complex brains like what we find in worms, arthropods, etc. A simple “node” or neuron in an organism may confer a slight advantage such as “learning” that a chemical is food vs. not food. More nodes cobbled up together may be able to confer greater advantages as more complex “learning” and cognition about the external world is incorporated and used to aid survival and reproduction, and so on. Look up “Evolution of Nervous Systems,” just for starters, there’s much more on the brain, morality, etc.
To william’s point, we see today many extant examples in the animal kingdom of very primitive brains all the way to highly complex brains and just about every stage of complexity in-between (just like we see just about every example from simple to complex eye structures). And the complexity of these brains correlates very closely with the complexity of cognition that the animals express. Is this a coincidence?
Are we to think that all the “simpler” types of brains, with “simpler” cognition are amenable to and allowed to evolve under NS, but that somehow there’s an artificial boundary where more “complex” cognition cannot evolve via NS? Where exactly is that boundary, and why does NS suddenly stop there? Because someone is incredulous or has an agenda of “magic-of-the-gaps” to proffer?
Did NS stop with the chimpanzees, our closest cousins, and suddenly “magic” kicked in for the human lineage exclusively? Or did NS perhaps stop producing “higher cognition” with the Australopithecenes, which after all looked a lot like chimpanzees early on and had similar brain sizes, and “magic” had to wait until the Neanderthals arrived on the scene? Or maybe NS stopped with the Neanderthals, and then “magic” kicked in to produce higher cognition in the Sapiens, even though we know from DNA evidence that the Sapiens interbred with the Neanderthals? Was there one specific generation when “magic” kicked in for the children but was absent for the parents? If so, did this happen to only one family or to many families simultaneously when the “magic-granting magician” decided to do so?
I could go on with this reductio, but you tell me where “incredulity” is warranted here: In NS working gradually over geological times to produce higher cognition, or in NS working gradually over geological times to produce “simple” cognition up to an unspecified, unexplained artificial boundary, and then requiring “magic” beyond that boundary?
LikeLiked by 2 people
Hi Eric,
I think we have reached a point in our discussion where we can just repeat ourselves again and again without making any progress. No matter which scenario you present (in this case a meal choice), I will ask which option you selected and why. If there is no reason for the choice I will say it was a random choice. If you give a reason for the choice I will point out that the reason implies a pre-determined cause and therefore does not require a supernatural uncaused cause interacting with our brains. Repeat ad infinitum. The reason for choosing one meal over another may be preference / taste, social impact, impulse to try something new, etc. and all of this is based on our brain, our predictions, our instincts and our memories of past experience and results.
I was not referring to evolutionary development. I was referring to our lifetimes and how we have all been exposed to different life experiences, different families and cultures, have read different books and articles, etc. and so we think differently because of this.
LikeLiked by 1 person
Hi Travis,
I have been trying to think through and read up a little on what you are saying. These ideas are just a tentative response to what you have been saying. I’m still thinking things through, but this is where I’ve got so far.
The first thing is I realised that your Socrates example is not directly related to natural selection, but takes place in an individual’s development.
Natural selection presumably led to the the setting up of the “fire-together, wire together” capability, and the tripling of brain size about 2 million years ago, both of which are necessary for human cognition. But I think the process you are using in your example happens as a person grows and establishes connections in their brain. And therefore it seems to me that:
(1) Your process works on specific examples, not on syllogisms in general. (If generalisation occurs, it must occur later.)
(2) It depends on the experiences people actually have. If a person doesn’t think of Socrates, they’ll never make the connection you describe.
(3) And if a person makes a faulty observation, e.g. by reading or hearing it, then a faulty connection will occur as we’ve discussed. If they get both good and bad information, they’ll have a fuzzy connection.
So I’m not sure it is logic or inference that is being learnt here, but simply empirical observation and memory. If the process occurs as you say, then the brain makes associations, but these associations are between the words “Socrates”, “Man” and “mortal” but not with “logic” or syllogism” or “inference”. So the brain has done logic but we don’t actually know it has done it. So I’m not sure that we have learnt anything about logic after we go through the process you describe.
And of course, if you have heard wrong things, your brain will have made different associations that may be erroneous.
So rather than learning logic, it seems we have learned to associate whatever we have heard, seen or thought about. Generalising to the logic of Modus Ponens is still awaiting our brain, and still requires an explanation. Maybe the same sort of explanation will work, maybe it won’t.
But this leads us to your question about why lower forms of cognition (like the Socrates example) may be possible if determinism is true, so why not higher forms?
Since we can see that how each brain develops depends in part on innate ability, and in part on what we think about and read and hear, then we will all clearly develop slightly differently. This will get worked out in common cognition because of common experiences, teaching and socialisation, but once we get to higher forms of cognition (and a long way before we get to Fermat’s Last Theorem!) there will be fewer shared experiences, more questions which don’t have easy answers, and therefore some very different brain processes. (Remember, we are still talking about cause-effect physical brain processes that have supposedly developed in ways that produce ground-consequence logic without actually doing the logic.)
I conclude then that, if determinism is true, we don’t do inference, our brains think cause-effect in ways that will be more likely to correlate or produce simple logic, but less likely for complex questions, with the outcome biased towards whatever inputs we happen to have received.
So, let’s think about what happens when two people disagree about some complex question. Person A presents an argument. Person B’s brain sets up some associations and compares them to the associations already there. If it is a contentious question which they already have an opinion on, the existing associations will be stronger and they will reject the argument, though maybe their commitment to their status quo will be slightly weaker. If they hear the argument enough, they may weaken so much that their view changes. So the argument works, not because it was “right” but because it overpowered the brain’s cause-effect processes by weight of words.
Is that inference? I don’t think it is. I think it describes what can happen to some people sometimes (it’s how much advertising works), but not what we experience when we work through tough questions. It doesn’t describe the discussion we are having.
I think choice raises problems for all viewpoints, but I still believe there are more problems for atheism/physicalism. Like the discussion with Dave, I’m not sure there is much more for either of us to say. What do you think?
LikeLike
Hi Dave,
You still seem to be arguing “choice” is a binary matter – either totally caused and determined or random. I don’t think you have presented any reason why it can’t be partly caused and partly random (e.g. some causes which can be resisted because they are not determinative, and then a non-caused – random? -choice is made between them) which could allow genuine choice.
I also don’t see why you mention “supernatural uncaused cause”, because I have been anywhere near arguing that. I have simply been arguing how determinism can produce genuine logical inference.
But I agree with you that further discussion isn’t likely to produce much more light nor any changes, so I’m happy to call it a day. Thanks for your time and ideas.
LikeLike
Eric,
You are differentiating between logical reasoning and inference by association, but I’m trying to suggest that they are effectively the same thing. We may learn to do formal logic, but I just see that as a system of rules we use to constrain and communicate the probabilistic, inferential process that is already present via the associative network of our brain. Can you give me an example of logical reasoning that cannot otherwise be understood as a process of inference from associations?
As for the second part about complex reasoning, I don’t think I fully follow your points but it does appear that your example overextrapolates from a literal reading of my simple example. The predictive networks incorporate a broad variety of inputs and concepts and just reinforcing one element doesn’t undo all the other associations that inform the prediction. I had meant for the premises in the simple example to be proxies for the vast array of evidences we receive to form our concepts and associations. That said, there is also evidence that we judge familiar things to be more true, so your characterization does play out to some degree. If anything, though, that’s just further support for the model, even if it does carry the potential to mislead. I’m not claiming that this framework makes us perfect reasoners, only that it gives ground for a general reliability according to the data we receive.
LikeLiked by 1 person
@UnkleE
Now, I want you to really, really, really try to apply some very serious critical thought to the next part of this question, okay?
It might be difficult for you, I realise , considering a worldview over much of your life based on fuzzy-connections , but please, try. Can you do this for me?
So let’s start. Apply the same reasoning to…
1.An omniscient deity that required a blood sacrifice …. his own kid, apparently, but really himself in human disguise, in order to forgive humans for what, we really don’t know.
2.The Virgin Birth of the Biblical Character, Jesus of Nazareth.
3.Dead people coming back to life and going Walkabout then disappearing up into Heaven.
4.A delusional, possibly epileptic, possibly even fictitious dastardly Christian-Hunter and self-styled apostle.
That is just four things to consider. Only four. Hardly anything, really. There is more info on your Cornflakes Packet.
Shouldn’t take too much critical thinking.
I would be very interested in hearing what you came up with after giving this some serious thought.
Thanks.
Peace and … stuff.
Ark.
LikeLiked by 1 person
Don’t overthink this topic of free will. Skewing the definition and eloquently stating theory and using convincing language and reason, doesn’t change the reality that, I chose to turn on WordPress this morning, I now I’m deciding whether to go to work today. I usually do, but today may be different. If I stay home that would be strange for me, and no one could’ve predicted it. I have free will to do, or not to do, and I am an atheist.
LikeLike
Hi Eric,
I’ve had that concept in the back of my mind from one of your first comments:
https://findingtruth.info/2017/06/06/difficult-questions-for-atheists-part-1/#comment-33733
“The only way then we could have free will is, as Travis says, there is some uncaused component in our thinking (I would prefer to say not totally caused by these processes). And a random process, if such exists, won’t help, because it isn’t under our choice either.” … “This means that the only way we can have “real” choice is for there to be something other than physical in the mix.”
I looked this up only to explain my line of thinking on this thread. I’ve basically been trying to wrap my head around the concept of something that is uncaused and yet not random. Anyway, I think we can safely jump off this carousel. Until next time! Cheers Mate! (Or whatever guys in Australia say)
LikeLiked by 1 person