Talk:Regress argument (epistemology)

merge?

Is there any reason this shouldn't be merged with Problem of the criterion? --Ryguasu 14:59, 3 Nov 2003 (UTC)

I think it is sufficiently different to leave as is - although the links between the articles need improvement...Banno 12:20, 3 Jan 2004 (UTC)

Critique

I think this article should stand on its own also, apart from Problem of the Criterion. I agree with Banno on that.

However, I do not think that this article does justice to the regress argument. Merely dismissing it as logical false seems a little one-sided. It is logically false only if one assumes the premise this author is trying to prove, i.e., that propositions of truth can be justified somehow. This refutation of the regress argument begs the question, by assuming the premise it tries to prove. In my opinion this is not a discussion of the regression argument, but a mere attempt to refute it in favor of foundationalism.

The regression argument seems more closely tied with skepticism than with foundationalism. It supports the skeptic more than the foundationalist, but the article states that it supports foundationalism. What happens when the regressional questioning is used to understand the basis of foundationalism? The foundationalist at some point says, “I no longer have to give any reasons for what I state. It just is.” To a skeptic this sounds like dogma and not philosophy.

The arguments against the regression argument and possible alternatives to it certainly should have their place on this page. But, it should be a page about skepticism, not about foundationalism. A refutation should not be the whole page. There is more than one view of the regress argument. The regression argument plays an important part of the dialectic that is philosophy. It should not be dismissed so quickly. The current article does not address it as objectively as it could be. -RyanKoppelman 1/3/04 14:31

I agree. After writing the Coherentism article I though a bit of editing for the sake of consistency was needed, but I have done the bare minimum. The article looks to me to be suffering from ‘too many cooks’ syndrome. Targeting scepticism instead of foundationalism sounds like an excellent idea provided the links are left in. Please, go ahead and re-write. Banno 22:10, 3 Jan 2004 (UTC)

"This article needs additional citations for verification." -pardon me for saying so, but in the context of a discussion of the regress argument that is extremely FUNNY ;-) —Preceding unsigned comment added by 76.107.26.117 (talk) 01:22, 22 January 2010 (UTC)

Edits to Merge

A re-arrangement as much as a re-write. I hope I haven’t left anything vital out – just add it if I have.

I’ve obviously changed my view with regard to the three articles. The main reason is that WikiProject Philosophy allows us to co-ordinate across multiple articles, but I also wanted to make the articles fit together with recent edits in epistemology - I’ve focused on the argument being a corollary of Theaetetus. I’ve also attempted not to let the argument lead directly to any point of view – one of the difficulties before was that one version argued for scepticism, another for foundationalism. As it stands, I hope one can see that all three are possible responses. Banno 23:52, Apr 10, 2004 (UTC)

Wrong from the first sentence

There is more than one Regress Argument, and none of them is identical to The Problem of the Criterion. (Arguments are not problems.) The Problem of the Criterion is a problem, a conundrum, not an argument. An article titled "regress argument" should be a general discussion of regress arguments, and the form of regress arguments--something that philosophers have written quite a bit about. --A philosopher —Preceding unsigned comment added by 69.107.24.167 (talk • contribs) 21:13, December 30, 2005 (UTC)

Reference to Carrol?

It's hard to imagine why nobody included a reference to Lewis Carrol's "Achillies and the Tortoise" short story, wich perfectly summs up the condorum. —Preceding unsigned comment added by 82.216.185.139 (talk • contribs) 13:14, August 3, 2006 (UTC)

That was my first thought as soon as I saw this article. Perhaps I'll add it, if I can get round to it. FrankP (talk) 13:41, 8 January 2020 (UTC)

Merge in "Infinite regress"

The suggestion was made several months ago go merge "Infinite regress" into this article. There has been no comment on either Talk page about it. Does that mean there is no support, and the tags should be removed? Mdotley 19:01, 6 October 2006 (UTC)

I'm a mere mortal here, but would suggest that they should stand apart. "Infinite regress" has currency apart from the more technical epistomological exploration contained in "Regress argument." It's meaning is more than a dictionary can bear, but in its wider use it would be buried by the magnitude of the "Regress argument" section. I chose Wikipedia over a dictionary because I wanted just what I got from the "Infinite regress" entry - nothing more ("Regress argument") and nothing less (a dictionary's less stringent approach). It was a good fit. Peace. -Hoover 24.155.252.87 04:55, 18 October 2006 (UTC)

Remove the Problem of the Criterion

The problem of the criterion is not the same as the regress argument. The problem of the criterion deals with how we know certain facts. For instance, how does one decide what makes is a "good apple"? Do we first identify good apples and then determine which characteristics it has? or, do we decide what the criteria are for being a good apple, and then decide which apples are good?

The problem arises when one comes to justify their selection of either of the methods. If we first select good apples, the test for good apples ultimately seems to be whether or not it meets the criteria. Yet if we choose the criteria first, it would seem that the test would be to actually select an individual apple as an example. This is the problem of the criterion, I'm really not sure why it is merged with the regress argument. —The preceding [[Wikipedia:Sign your posts on talk pages 24.92.249.43 23:32, 14 February 2007 (UTC)|unsigned]] comment was added by 24.92.249.43 (talk) 23:31, 14 February 2007 (UTC).

Greetings. I actually created a wiki account for the first time to resolve that issue; I punched in 'problem of the criterion' only to be redirected to regress argument. I made a kind of make shift problem of the criterion page since then which hopefully will get cached out by someone more wiki-enthused then myself. This article is rather odd in itself. The regress argument is a problem that occurs within theories of knowledge that define knowledge to have justification for belief as a requisite. If all beliefs must be justified, even the justifying beliefs, then either the chain of justifcation ends with an unjustified belief, a circular justification (X is justified by Y, Y is justified by Z, Z is justified by Y) or an infinte regress of justification. I'm not entirely convinced that this should be an article in itself, seems to be a topic within justification. McDuderson (talk) 13:27, 18 May 2008 (UTC)

What does it mean for something to be "justified?"

This section reads argumentatively. It also does not really solve the regress problem. Proof proceeds from axioms. But what proves the axioms?

1Z 10:47, 20 August 2007 (UTC)

Good point. Actually, the article has become a bit of a mess. Lots of Wikirot. Needs a good re-working. Banno 12:16, 20 August 2007 (UTC)

From the article:

That this is not a stupid question becomes apparent when we exchange "justify" for "proof". On it's own the word "proof" means nothing. The notion of proof is inextricably tied to assumptions, and the same goes for justification. The reason this might not be obvious is that in everyday usage the assumptions being used are typically dropped. But that is not okay when a precise discussion is required.

Noting the above, and restricting ourselves for the moment to mathematical considerations, there are two apparent possibilities. We can either mean "proof from the axioms of <some formal system, for example PA>", or we can mean "proof from intuitively obvious principles". It seems that every use of the word "proof" in mathematics would be covered by the above two meanings, and there is no other coherent meaning one could give the word "proof". Then, under the first meaning, which is the one used in formal logic, some true statements can be proved and some true statements cannot be proved (some theorems of number theory), and under the second meaning, which is the one used in mathematics (with the exception of formal logic), some true statements can be proved and there is no reason to suspect that there are any true statements which cannot be proved.

Now in either case, we have that some true statements can be proved. How can this be? Surely this has not happened merely because we stopped using the word "justify" and started using the word "prove"? The real reason this has happened is because we explicated the assumptions being used. The old regress argument was technically meaningless owing to an insuffciently precise notion of "justification".

One could then say "proof of A" means "proof from intuitively obvious principles, none of which is A, and none of which is equivalent to A". Then, it is well known that some true theorems cannot be proven, such as the principle of induction; any proof of the principle of induction must necessarily use the principle of induction. But in this meaning of proof, acceptable grounds for proving some statement depends on the statement. If we require that: "if certain principles justify some statement A, and these principles can be used to prove some statement B, then B is justified" then the above is not an acceptable meaning of "proof".

Editing for Bias/Other Problems

"Assuming that knowledge is justified true belief..." Why not just say, "Assuming that knowledge requires justification..."? I think it would get the point across just as well and it wouldn't narrow our view of knowledge down to justified true belief.

"The most common objection to naïve Coherentism..." I think this is inherent bias. I'm not a Wikiexpert, but perhaps this should be tagged?

"...and thus seems to prove that Coherentism is not true because beliefs can be justified by concepts other than beliefs." First, I don't think this is correct. We haven't required that knowledge be required for justification. Our belief in the white canopy bed is justified, but it may/may not be knowledge. This passage is especially confusing if one doesn't already know what the Regress Argument is. We haven't listed in our premises that concepts must be justified by beliefs. Here is my suggestion for reworking:

Let us include the following premises for the Regress Argument:

1. Some beliefs are justified
2. If a belief is justified, then it is justified by another belief
3. No belief can justify itself, either directly or indirectly
4. No belief can be justified by a non-terminating, non-repeating chain of other beliefs
5. No belief can justify another unless it is itself justified

From this list, one may deny any of the premises for the following results: deny 1 (Skepticism), deny 2 (Foundationalism), deny 3 (Coherantism), deny 4 (Infitism), deny 5 (Positism), deny a combination of the premises, or deny none of the premises (paradox).

I didn't want to make any changes myself firstly because I don't have the energy to rework the entire page, and secondly because I wanted to post under the discussion first and see if people had any opinions. Jadecoppieters (talk) 02:20, 12 February 2013 (UTC)

Is diallelus correct?

The very start includes "also known as the diallelus." I'm not a philosopher, but I dabble for intellectual recreation and in my travels "diallelus" has always been specific to circular argument as distinct from regress argument. I followed through the citations for that paragraph and none of them record diallelus as an "also known as." Indeed they are pretty clear that circular argument is not regress, but circular argument is sometimes used as a counter to regress.

I may be reading it all wrong, but if not it seems significant in the first sentence.

Caleval (talk) 00:18, 6 July 2017 (UTC)

I added a source for the term. If you have sources making a distinction between the two, please list them. This article is sorely in need of references. Paradoctor (talk) 21:06, 6 July 2017 (UTC)

Merge request

I propose that this article be merged with Münchhausen trilemma, one way or the other, though I prefer merging this article there. These articles really describe the same problem - that any proposition can be endlessly questioned as any justifying statement must be justified as well, and that there are only three options for conclusively ending the argument. --37KZ (talk) 22:56, 2 January 2020 (UTC)

• Oppose The current title of "regress argument" is reasonably clear as plain English. The term Münchhausen trilemma is less clear, being a fanciful, fictional analogy. That title is something of a neologism and it's not clear that it is the common name for the topic. For example, the Stanford Encyclopedia of Philosophy has a comparable entry of Infinite Regress Arguments while it doesn't even seem to mention Munchhausen. There's some talk there of Agrippa's Trilemma and that seems a better title as it alludes to a classical philosopher, not a tall tale. But Agrippa had five modes and so it's not clear that we should restrict coverage to just three of them. Andrew🐉(talk) 23:32, 2 January 2020 (UTC)

It’s still the same thing, so we could mention both titles in the same article. --37KZ (talk) 23:42, 2 January 2020 (UTC)

• Oppose for same reasons. Teishin (talk) 00:51, 3 January 2020 (UTC)
• Oppose They are obviously linked, but neither topic seems to comfortably contain the other. The trilemma includes regressive arguments as one of its horns, but the regress argument is sufficiently involved to require an article of its own. So each should make reference to the other, but they stand alone quite naturally. FrankP (talk) 13:43, 8 January 2020 (UTC)
• Oppose, although this article needs work. Banno (talk) 16:36, 25 January 2020 (UTC)

Primus

@AndTitusSaid: You cite a paper that has been very recently published. It is unclear if this is some sort of consensus, held by many scholars in the field or if this is an opinion which was widely rejected or even ignored. Thus it lends WP:UNDUE weight to this paper in an article which briefly describes the major schools of thought on this matter. Also, the use of HTML-tags in articles is deprecated. Kleuske (talk) 18:16, 11 April 2020 (UTC)

Something is missing

How is "any proposition requires a justification" itself is justified? 88.115.35.152 (talk) 22:46, 27 April 2020 (UTC)

Godel's incompleteness theorem

Being a scientist and just stumbling upon this article it feels like it should reference Gödel's_incompleteness_theorems. I am not familiar enough with either the theorem and the regress argument to add the change myself, but as I understand it Gödel is actually saying that any attempt to prove everything (answer all the questions in the regress argument) leads to contradiction (especially with this type of enumerable arguments), and thus any complete theory is either inconsistent or must be based on axioms (unprovable statements). This really feels like a form of answer to the regress argument. Is there anybody familiar enough with both to confirm my hunch? Dragomang87 (talk) 18:45, 9 November 2021 (UTC)

It could be that such a connection can be drawn. But we would need reliable sources that draw this connection to avoid original research. The Stanford article on Infinite Regress Arguments does not mention Gödel, which would indicate that this topic is not of central importance here. Phlsph7 (talk) 19:19, 9 November 2021 (UTC)