Does one need faith that a ball will drop to the ground because they’ve deduced that it will through inductive reasoning? The problem of induction says that past results can’t be relied upon to be accurate, so it must take some sort of faith, no?
Not really. Common sense would go that “what works, works.” If people had to prove everything, we would have all wasted away millenia ago.
Faith is belief based on a lack of proof. Inductive reasoning is making an assumption based on experience. But just because you are making an assumption doesn’t mean that you “believe” that a certain outcome is necessary. It just means that you think the probability is great enough that it’s not worth bickering over minutiae.
So given that you don’t “believe”, and you do have some amount of proof, it’s automatically not faith.
I think the answer to the OP’s question depends on what he means by “faith.”
Inasmuch as every line of reasoning requires premises* and inductive reasoning itself makes presumptions (i.e. causation’s repeatability) , then yes, faith is necessary in making deductive reasoning.
*Take faith in your senses as an obvious example.
Technically speaking, that’s inductive reasoning, not faith.
If you take a definiton of faith that’s synonymous with “inductive reasoning” then, yeah, one is the requisite of the other. And I agree that that’s a popular meaning for the word, but also sort of pointless to debate (“they’re the same”), so I assumed the technical meaning of the word rather than the vernacular.
Well, inductive reasoning requires belief in the claim that the future will be like the past, which is something we can’t prove in a non-circular fashion. What would such a proof look like? “We know the future will be like the past, because in the past the future has always been like the past, so we can infer that the future will, as it has always been, be the same as the past.” Reasoning doesn’t get any worse than that. But I think this is just a special case of the general principle that no method of reasoning can be self-vindicating: there is no non-circular way to prove the reliability of reason, or sense perception, or…
Science assumes that while as the constant numbers used in physics math might change, the formulas stay the same regardless of anything. If you’re willing to make that assumption, then assuming a future based on things that we have determined the formulas for, stops being inductive reasoning. For instance, “For every action, there is an equal and opposite reaction.” This can be viewed as fact rather than experience.
If you’re really hard core on that nothing can be proven absolutely, then you pretty much have to relegate all of science to relying on inductive reasoning.
Unless I’m way off base (it wouldn’t be the first time), inductive reasoning is very simply the extrapolation of a pattern from a limited set of data; hence, it assumes some meaningfully accurate and consistent way of acquiring data – in the example given by WamBam, the human senses. Since humans are rather limited to their senses, which we know to not always be accurate or consistent (at least in certain ways), the presumption that they are in any given case is a matter of faith. I will certainly admit, though, that we could just leave it with the descriptor “presumption”.
And now I wait for morning for the GD regulars to pummel me into the ground, as I doubtless deserve.
Varying data still has a general pattern and margin of error, so your assumptions are based on something–even if you might be making a poor judgement choice by assuming a smaller margin of error than is accurate. Faith lacks any evidence.
First of all, that’s a pretty narrow definition of faith. (Granted, the opposite extreme, “lacks absolute proof”, would be going too far.)
Secondly, any sort of self-referential calculation of a margin of error is a matter of memory, which is really tricky.
But to be fair, “presumption” really is a better word, seeing as there are a lack of other things to believe. At the end of the day, WamBam, you have to have premises. Unless you feel like imitating Descartes (and, IIRC, he more or less gave up his line of reasoning anyway), “faith” as we talk of it just uses complicated premises, instead of simple.
I don’t think “faith” is the right word. It’s far too loaded. In some sense induction is merely deduction with an extra premise. In a similar sense deduction is merely an interesting form of induction where all propositions are restricted to either 100% or 0% truth values. Given that deduction and induction both spring ultimately from the same source (human intuition), it doesn’t bother me that we cannot deduce induction from deduction. Especially since we can induce deduction from induction. 
My own opinion on the excellent point that Sophistry and Illusion brings up is that there is no need to prove that the future is like the past. There is simply no reason to assume a priori that the future and past are two separate things.
…when, in fact, they are many separate things, n’est-ce pas?
However we need make no assumption about the consistency or accuracy of our senses. Since inductive reasoning is not guaranteed to give correct answers, we must always check the results against reality. Doing this has shown how inaccurate our senses are (how else would we know this) so we would much rather rely on our instruments. What’s more accurate in measuring your speed - your senses or your speedometer? So we don’t need faith at all in it, since we are continually checking our results, and are aware we may be in error.
Someone who is sure their senses are always accurate would be working on faith, but that person would be using induction incorrectly.
This is an interesting move. I have read a bit of philosophy of time, and the dominant view (which is apparently heavily influenced by the view of time advocated by modern physics) is that there is no significant ontological difference between past, present and future. From the Stanford Encyclopedia of Philosophy entry on time:
I haven’t heard this proposed as a solution to the problem of induction, but it definitely seems worth looking into.
And what apparatus does the testing mechanism employ? One’s senses.
Let me illustrate. You wish to check if table A is longer than table B. You have a look and decide that table A is noticeably longer. Next day, you wish to confirm your result. You have a look and now the tables appear the same. Assume that, ontologically, no physical change has been effected upon the tables. Baffled by the discrepancy, you fetch a tape measure and decide to obtain “objective” numbers to settle the matter once and for all.
Question 1: what process will you use to obtain the numbers? In the end, you’ll look at the tape measure markings.
Question 2: Re: accuracy, are you sure that your sight is properly interpreting the visual signal corresponding to the sight of the markings?
Question 3: Re: consistency, are you sure your faculty of vision is consistent across the two observations (table A marking and table B’s)?
Clearly, the past and future are separate things. There’s Change, which is how there’s experience of Time at all, and a position on its axis i.e. noon, evening.
The problem, as I see it. is that due to the notion of other subjects existing, there’s this implicit metaphysic of an objective reality i.e. some reality constant across subjects. And since the world is consumed only via the senses, the quest to arrive at this objective reality is naturally, via the senses, the very apparatus which is subject-mediated. Our existence is epistemologically solipsistic (if that’s not true for you, even then, especially then, it is true for me) and hence any effort to discover objective reality is bound in uncertainty. Adopting a functionalist stance is more parsimonious than the ‘objective’ reality. Incidentally, adopting such a stance precludes questions like "But, what’s really the case?’.
Nonetheless, let’s assume that the past and future are “the same”. What does that mean? Clearly (to me), it does not mean that my toothpaste displays the same shape upon the regular periodic inspection of 15-16 hours. That aspect exhibits change and hence the ‘Past Shape’ is different than ‘Future Shape’. So, there are two possibilities left,
a)some observations are constant across all time, and some aren’t,
b)no observations are constant across all time.
In this formulation, the framework shifts from induction i.e. a temporal concept to classification of observation i.e. a taxonomic concept
Instead to asking
a)Will the law of “like charges repel” hold tomorrow?
We can ask
a)Is the observation “like charges repel” indeed a law, wherein the concept ‘law’ applies to only eternally constant observations?
And the more parsimonious answer, to me, is No. If Change is the most fundamental characteristic of sentient existence, then that ought to include all “laws” as well. There may be no “ground”. It’s turtles all the way down, and the only constant is change, and that includes the “rate” of change and other calculus-inspired metafunctions of change.
We can do the following experiment.
Get N people (or one person making N measurements separated in time) to state if the table is longer or to estimate the size.
We can do the same thing by having them measure the table with a tape measure.
We can analyze the variance of the results. I’d predict that the variance of the estimation is much higher than that of the tape measure measurement. Repeat for as large a sample size as you like, and you will no doubt conclude that the tape measure is better than the eye.
Now, is your example that of the well know optical illusion, making two same sized objects appear to be different sizes? We might note interesting characteristics of the two tables. We might remeasure with different tape measures, and get the same results. Then, we might bring in two tables of different sizes, but more standard, and note that observers now get it right. Based on this we’d no doubt commence a set of experiments to help us understand why our eyes got it wrong for the first case, and right for the second.
It’s certainly possible for there to be outlier measurements. One of the N people reading the tape measure might exhibit wide variance. The next step would be to understand why - perhaps you note he has a guide dog with him.
The point is, the reason we rely on instruments is that they exhibit less variance and now, since they can output measurements in a variety of formats, they are repeatable. (Not taking the measurement, but reading it.) In the old days if you saw something interesting on a scope, you photographed it, which was a pain. Today, it can be easily captured and emailed. It’s all measurable.
Rather frivolous hijack: This discussion is the reason I’ve always hated that saying about assuming. (“You know what happens when you assume!” Said in that annoying sing-song) On some level, everything you know is an assumption (i.e. involves inductive reasoning). You can’t *not *assume.
I think Gyan’s point is that, in order to read the instruments and know what the measurements are, we still have to use our senses. We can’t leave our senses completely out of the process; we have to use (and, therefore, rely on) them at some point.
If every time I observe an unsupported rock, it falls, there is a fairly good empiric reason to plan my life around the premise that unsupported rocks fall. Generalizing that to the case that unsupported stuff in general falls is almost as reliable. Both cases involve incorrect assumptions, and false logic. But most of the time, it does work. Helium balloons, orbiting mind control lasers, and other such things do force those who care to investigate their underlying assumptions.
Now, if you want to find out why stuff falls, and why the stuff that doesn’t fall doesn’t fall, you have to tighten up your thinking by an order of magnitude or so. The overwhelming majority of humans in history mostly don’t give a damn.
Tris
The race is not always to the swift, nor the battle to the strong, but it is the way the smart money goes down.
And how do you receive their statements? Primarily via your visual and auditory faculties.