One man's look at the debate format in Wikiversity

This article describes Dan Polansky's vision for debates as currently implemented in Wikidebate, which can be equally well implemented in other public Internet place and as part of personal or corporate decision making. One can see these debates as an analogue of a British debate as seen in some YouTube videos. However, one can also see them as a tool for argument analysis and public collective cognitive psychotherapy. The tool can be useful also for a person using the tool in private offline as a means of breaking out of dogmatic uncritically accepted beliefs and misconceptions. One could also use the tool in Wikipedia as a complement for requests for comment; as long as a request for comments focuses on arguments rather than numerical vote counts, one should be able to construct a debate corresponding to the request that abstracts away from the debate participants, like a debate in the format of Wikidebate.

Debate page features
Features of this kind of a debate page:
 * 1) There is a yes/no motion formulated as a question.
 * 2) There is a heading stating the motion in affirmative.
 * 3) There is a section for arguments for.
 * 4) There is a section for arguments against.
 * 5) Each argument can have objections as subordinate nodes, whether for argument for or against the motion.
 * 6) Each objection can have objections as subordinate nodes, and then recursively.
 * 7) Arguments and objections are not signed (unsigned). They do not represent speakers. One cannot tell how many speakers there are; e.g. different arguments for can be from a single speaker or multiple speakers.
 * 8) Questions are not allowed, upon strict interpretation, unless one finds a way how to interpret a question as an argument.

Lack of signature
The lack of signature on the items (arguments and objections) fosters the idea that we are analyzing arguments in disregard of persons, which helps avoid the ad hominem logical fallacy. Moreover, it contributes to the idea that one person can formulate the argument while another person can improve the argument, whether by changing the formulation, adding references or splitting the argument into two. Furthermore, it facilitates the idea that an editor should only add an item if the item is not already covered by an existing item, which is in stark contrast to how Wikipedia conducts its requests for comment, which often look like votes/echo chambers.

The support relation
The debate format crucially depends on the support relation. That is, a statement is meant to support another statement. Thus, an argument for supports the motion, whereas an objection to the argument supports the negation of the argument or points out a flaw in it.

What this support relation consists in is rather unclear. What it does not do is provide a strict logical inference. This relation is probably studied in argument theory, which is not a branch of formal logic.

Argument and objection moderation
A key question is how to moderate arguments and objections. Some principles:
 * 1) Each item needs to meet some minimum standard of relevance. The standard is left unspecified.
 * 2) Each item needs to meet some minimum standard of convincingness. Thus, there must be a non-trivial number of readers who could be convinced that the item is valid. Unfortunately, this is not objectively measurable.
 * 3) An item does not need to be correct, merely somewhat convincing. Indeed, if arguments were supposed to be correct, there would be no need for room for objections.
 * 4) Vulgarities can be censored: any argument that can be made using a vulgarity can also be made without it.
 * 5) Sometimes it is better to write a rather strong argument than writing a weak argument and then allowing an objection to it. However, sometimes it can be worthwhile to see how an initially weak argument is made, an objection is raised, and a stronger argument derived from the weak one to address the objection is presented as a separate item. This can be rather educational.

Limits on questions asked
Some principles for the types of questions allowed as motions:
 * 1) The question can be a matter of descriptive fact or of value. For instance, the question of whether God exists is one of descriptive fact, albeit unknown whether it is true upon strict logical analysis. The question of whether metaphors are good is probably one of a descriptive fact since "good" is used in an instrumental sense. The question of whether slavery is good is probably about a mixture of instrumental good and moral good. (One may also see moral good as a species of instrumental good, that is, good for morality.)
 * 2) The question can be about a value or norm that the present Western society finds unacceptable. The debate does not endorse the motion and it does not endorse the arguments made. For instance, there can be a debate about whether slavery is good, complementing Wikipedia's article on Proslavery thought.
 * 3) The question has to be minimally interesting and controversial at least in some context or in some time period. For instance, whether slavery is good was actually debated in America. For another one, whether the Earth is flat is debated on YouTube, and one may find oneself amused by reading the arguments made by the flat earthers in a debate format, or one may be inspired by the calm and thoughtful way in which someone responds to arguments that appear absurd. Having such a debate does not endorse the notion that the Earth is in fact flat. Indeed, e.g. Sabine Hossenfelder's consideration of the flat Earth theory is not her endorsement of the theory.
 * 4) The question can be somewhat vague or ambiguous so that the argument has certain freedom of interpreting the words in the question, especially "good" and "should". In a technical analysis of pros and cons, "good" and "should" need to be disambiguated, often reduced to technical evaluation criteria. These word can refer to ethics, expediency, instrumentality, etc. Different pro and con arguments can refer to different evaluation criteria; it is up to the reader to choose which criteria they consider most salient or convincing.

Censorship
On the whole, censorship does not seem to be a good idea. Nor is the idea that a debate on an objectionable motion is incitement to violence, harm or discrimination.

Someone could argue like this: There should be no debate Should abortion be legal? since this debate argues for or incites to a murder of innocent members of human species. And yet, anti-abortionists are not using this argument to stifle debate; rather, they are openly debating the opposition. Britannica's procon.org features arguments for and against at https://abortion.procon.org/. Granted, this question is currently not settled and is subject to a public debate, unlike, e.g., whether slavery is good. But then, the morally sensitive topic's not being settled may make the debate even more dangerous: slavery was firmly rejected by the U.S. and Europe, while deviously clever arguments for the wrong side of the abortion debate (wrong according to one's conviction) run the risk of materially contributing to the wrong side winning.

Moreover, there are much graver concerns than slavery, especially when one looks outside of one's speciesist (human-centered) ethical bubble and instead considers other species or portions of biosphere as worth more than zero intrinsic value. Thus, there is the concern with species extinction and what to do about it. There is the risk of global climatic change, and debates to be had about to what extent there is a change, to what extent it is caused by human activity, whether the models are reliable enough to be taken seriously, what is the role of precautionary principle, should we "stop oil" as soon as possible, what is the possibly corrupting role of research interest (being a funded researcher) in arriving at truth but also of corporate interests to sponsor fringe scientists to create an appearance of a debate, etc. It seems unwise to try to resolve these grave concerns by stifling debate.

Related to this are sections Dangers and concerns and Platforming.

Dangers and concerns
Some dangers of or concerns about the debate format:
 * 1) The format suggests relativism. It suggests that for every motion, there are arguments for and against and no ultimate decision can ever be reached. This objection has some force but is unconvincing upon a closer analysis.
 * 2) The format allows an accumulation of unpalatable subjects and arguments. However, that is what often happens in serious philosophy. For instance, Popper's Open Society, volume 2, surveys unpalatable views of certain German philosophers on war. In a sense, Popper platforms these philosophers, accumulating some of their output in his volume. Popper vehemently disagrees with Hegel and Marx, and yet, what his writing does is giving the two thinkers a platform to some extent, accompanied by criticism. In a wikidebate, arguments are not left without criticism but rather accompanied by objections.
 * 3) Since the relation of items does not depend on strict logical inference or proof, one can entrap oneself in a debate and end up thinking something untrue is true or something unjust is just. However, it is not obvious that the generally accepted format of a philosophical monologue or treatise fares any better. Indeed, a contention is that the monologue often fares much worse than a debate, locking the reader in a certain narrow dogmatic view, without having the reader consider the possible objections to what they read.

Disclaimer
To prevent misunderstanding, one can add a disclaimer to a debate to the effect that Wikiversity does not endorse the motion or the arguments, and explain the value of having such a debate anyway.

For instance, Is slavery good? features the following:
 * Disclaimer: The arguments for the motion do not represent the view of Wikiversity. Wikiversity editors do not assert that slavery is good, just, morally acceptable or that some people are slaves by nature. The purpose of this page is to examine arguments in a debate format, including arguments one disagrees with.

Opportunities
Debates offer the following opportunities:
 * 1) Provide an analogue of British-style debates available in YouTube, albeit in a textual form. Unlike a video, the text provides a great searchability and copy-and-pastability.
 * 2) Act as a fact-checker and refuter for often claimed misconceptions and mistaken arguments. In order to be countered or refuted, the mistaken argument or statement needs to be stated and thus platformed.
 * 3) Teach people how the interaction between arguments, counterarguments and objections work on an interesting set of examples.
 * 4) Serve as a data basis for an empirical investigation of fallacies people actually make when arguing about various subjects. In this sense, a debate is not an authoritative account of anything other than some people's foolishness.
 * 5) Provide an argument aggregation and analysis tool for debates related to Wikipedia and related projects.
 * 6) When used offline, make sure that possible objections and counterarguments to one's positions and arguments are explicitly recognized/stated, even if one originally thinks they are invalid. One can learn to try to see the matter from an angle different from one's own. One can then figure out how to best formulate objections to items that one considers invalid.

Platforming
One risk worth a more detailed consideration is that debates create an opportunity to host/platform very bad ideas, whether descriptively factually wrong or morally wrong. However:
 * 1) Those who are interested in those ideas are more likely to find them outside of Wikiversity anyway. The added harm does not seem considerable.
 * 2) The attempts to deplatform and censor ideas considered wrong are all too likely to contribute to conspiratory thinking. If the powers that be are so afraid of an idea or statement that they must censor it and prevent debate, it is only rational to think that that at least weakly suggests there is something wrong with the powers that be, and that they are at least to an extent engaging in quasi-conspiracy that they want to protect from being discovered.
 * 3) In a Wikiversity debate, each statement is open to a list of itemized objections directly under that statement. That is not so in the conspiracy theorist's favorite website where the same problematic statements can be found, albeit uncountered. This creates an opportunity for Wikiversity to find best objections and counterarguments to bad ideas that one can figure out, de facto acting as a fact-checker. Use of Wikipedia-style references to authoritative/reliable sources is available.
 * 4) One may be frustrated with having to explain about very unreasonable ideas why they are wrong. But that does not make it impossible in general.  At a minimum, an objection can state "That is wildly incorrect" or "That is a wild conspiracy theory" if one believes such an argument has force. A more specific response sometimes applicable is "That is an empirical hypothesis and as such it requires proper sourcing." Another one is this: "It is not obvious and therefore needs to be proved rather than assumed."
 * 5) Paradoxically, sometimes it is worthwhile to catalog absurd objectionable ideas without even trying to refute them. This is the case of Fact Sheet on the Elements of Anti-Semitic Discourse, which catalogs some bizarre anti-Semitic ideas, at least bizarre to a modern mind.

As pointed out in Disclaimer, one can mitigate the platforming concerns by placing appropriate disclaimers. However, one can object that such disclaimers merely pay a lip service and the objectionable content stays. It seems that objectionable statements are taken to be some kind of virus of the mind against which well-formulated objections are powerless. Even if it is so--and it is not obvious and we do not seem to know--Wikiversity is unlikely to contribute to spread of ideas not available elsewhere, while it does create an opportunity for attempted refutation.

Examples in literature
Some examples of a philosophical dialogue, although not constrained to a strict debate format:
 * 1) Plato's dialogues.
 * 2) Galileo's Dialogues Concerning Two Chief World Systems
 * 3) Hofstadter's Gödel, Escher, Bach
 * 4) Lakatos' Proofs and Refutations.
 * 5) In Popper's writing, there is at least one dialogue. Even the monologue writing is often something like an implied dialog, where positions of philosophers are stated and then criticized.
 * 6) Darwin's Origin of Species has a chapter on possible objections. Thus, it is as if Darwin is debating detractors, even if they remain nameless and the format is one of a monologue.
 * 7) Black's The Identity of Indiscernibles

A reservation is that the above items are rather different from the strict debate format and are not a result of collective wiki editing.

Implied dialog
What one can often find in philosophy is an implied dialog or its elements. It may take the following form: "One can argue that X. However, Y." Philosophers, rather than talking to themselves or leading an isolated monologue, rather often refer to ideas of other philosophers and make comments on them, whether they name the philosopher or not. While the explicit strict argument-counterargument-objection format seems rather unusual, elements of the claim-counterclaim structure embedded in monologue are not.

Last word
An argument or objection (item) with no further objections may superficially appear to have won and been accepted. But that is not so. The lack of objection is ambiguously interpreted as:
 * that the item is plausible and no objections come to mind or
 * that the item is so implausible that it does not deserve a response or that the responding party gave up to terminate the otherwise interminable debate.

Interpreting the debate is up to each reader. It is for the reader to accept or reject the items in the debate, based on an assessment that at least to an extent is not captured in the debate. Thus, a debate is typically an incomplete capture of a dialectic cognitive process; part of the process is left without articulation, only in the mind of the debate authors and readers.

Plaintiff, defendant and judge
One inspiration for the debate format is the model of plaintiff, defendant and a judge. A key idea is that the opposition is allowed to speak and find the best formulations of a defense possible. One point of contrast is that there is no judge in a Wikidebate.

Socratic dialogue
While both the Socratic dialogue and Wikidebate format appear to be dialogues, there are marked differences:
 * Asking questions is a key element in the Socratic dialogue. It is not so in Wikidebate, where, upon strict interpretation, questions are not allowed.
 * There are persons in the Socratic dialogue. By contrast, Wikidebate abstracts away from persons speaking isolating a reasonably atomic (non-compound) argument or statement (objection) as the central entity type.
 * Items in the Socratic dialogue can be relatively long and compound, whereas Wikidebate aims at having the items rather atomic. In its atomicity, Wikidebate may bear some resemblance to Wittgenstein's Tractatus.

Links:
 * Socratic dialogue
 * Socratic method

Engineering review
One inspiration for the debate is an engineering review of a textual object, especially one that is not a source code, including specification and design documents and their items such as individual high-level or low-level software requirements. When someone brings a comment/defect on that review, the author has to either accept the comment or explain why the comment does not apply. Somewhat similarly, when an opposer of a motion makes a counter-argument, the supporter either has to explain that it is invalid by means of an objection or make it seem the counter-argument prevailed. What an author of an engineering artifact cannot do is remove a comment/defect from a review (censoring it); the comment/defect becomes part of a record.

Some differences:
 * On an engineering review, someone has an authority to make a decision in case the author and the inspector/reviewer raising a comment cannot come to an agreement via a discussion. In a Wikidebate, there is no one to act as an authority to resolve disagreements.
 * On an engineering review, a comment is part of the record no matter how bad or irrelevant it is. By contrast, moderators of a Wikidebate should probably allowed to remove too bad/irrelevant comments. It does not seem acceptable to allow any misguided comment (arguments/objection) made by an IP in a Wikidebate and feel required to let it stay and feel compelled to carefully explain why the item is very bad and not worth allowing into the debate.

Argue for opposition, then for yourself
Based on memory, Stephen Covey's The Seven Habits of Highly Effective People mentions an idea that would be sourced from a Roman philosopher or other kind of notable: a good lawyer first writes the best speech they know how for the opposition, and only then for themselves. The debate format forces one to take the opposition seriously, by allowing at least the most persuasive--even if wrong--arguments into an express record. The putative lawyer can then try to formulate best objections they can find as responses to the arguments of the opposition.

One interesting feature of this is that a skilled arguer can often formulate the arguments of the opposition better than the opposition itself. A downside of this is that the skilled arguer can force themselves into the discomfort of realizing that the opposition has a much stronger case than the arguer originally thought; in the "worst" case, the arguer has to switch sides (which, if the arguer is after truth, is a good outcome, for truth at least if not for the arguer).

Epistemic authoritarianism
The debate format abandons various forms of epistemic (getting-to-know) authoritarianism. Thus, a Wikipedia article is expected to be so well verified and curated that each sentence not edited away can reasonably well be believed, unless it is embedded in a debate-like or comment-like manner in another sentence or paragraph. (Of course, the article may not have been well reviewed, may lack tracing to sources, and the sentence may just have been inserted by a vandal, but the curation idea is there in principle.) By contrast, a debate is open for sentences that are not worthy of believing, and their not being so worthy is documented in the objections to them. Wikiversity editors do not as if "sign" the sentences in a debate (do not claim the sentences have been checked for accuracy and argument strength) and do not tell the reader which of the sentences are right. Thus, many a reader has to exercise a skill they have only imperfectly developed: critical reading of sentences that appear to be plausible but are in fact defective. On the other hand, Wikiversity editors help the reader find possible objections by articulating them. Still, there is no guarantee that a skilled editor provided the necessary objections by the time the reader got to read the debate.

Proofs and Refutations
The book Proofs and Refutations by Imre Lakatos presents the idea that increasing something like an expository surface for refutation is an element of proof. This may be not the exact formulation, but should be close. Increasing of the surface consists in adding statements/formulations in which one may try to find an error. Thus, if a mathematician supports a theorem by stating a lemma on which it depends, this does not present a complete proof and yet is a proof-like step since now the reader can try not only to refute the formulation of the theorem (or conjecture, before it has been proven) but also the formulation of the lemma. Perhaps the act of supplying inconclusive material in support and opposition acts a bit like that, increasing the amount of material that the reader can criticize. However, this consideration is very tentative and inconclusive.

Skepticism concerning debates
There are valid (not outright dismissable) concerns about the soundness of the debate format, some of which are raised in Are wikidebates a good thing?. A skeptic or pessimist can charge that the format is fundamentally unsuitable for sound academic work and for sound acquisition of knowledge. See also section Dangers and concerns.

If we assume that such skepticism is right (an if), it does not necessary mean that debates should be removed from Wikiversity. It may mean they need a proper disclaimer and warning instead. The debates can still serve as a tool of exploration of the dialectic processes of individual human minds, small human groups, large human groups, cultures, etc. Since, many people lead something like debates in their minds concerning various subjects, and many people debate subjects between themselves. Debates are broadcasted over radio, television and YouTube; Western societies have not yet reached the conclusion that a debate format is fundamentally bad and ought to be abandoned. Thus, debate-like cognitive and decision making processes are operational on multiple levels in human societies whether we like it or not, and we can learn something about these processes by exposing and documenting some of the candidate specific elements/building blocks from which they are built, specific for each topic under debate.

Single-author debate
The debate format used by Wikidebates can be used as a format for argument analysis by a single author, an alternative to the treatise format, which is one of a monologue. The author becomes in part something like a devil's advocate by trying to figure out and discover arguments against the position that they initially find the more convincing one, as if someone was paying them to defeat their favorite side. The author would usually ask the following questions:
 * What reasonably relevant and interesting arguments for can be found in sources online?
 * What reasonably relevant and interesting arguments against can be found in sources online?
 * What reasonably relevant and interesting arguments for come to my mind, whether based on my own imagination or based on what I have read or heard?
 * What reasonably relevant and interesting arguments against come to my mind, whether based on my own imagination or based on what I have read or heard?
 * Which terminological quandaries and ambiguities are present in the words and phrases used in the question?
 * What objections can I discover using external sources and my own imagination, as well as arguments that I must have read or heard but do not remember where?
 * What is some of the best further reading I can find concerning the question asked?
 * Is there some further reading that has the form of a debate or at least arguments for and against, such as a procon.org page?

Icons
Currently, classes of items in the debate are marked using icons: While the icons do help somewhat to get an overall impression of the meaning of the debate parts, they are not essential. If someone finds them problematic, they can be removed with little loss of value, if any at all.
 * A green plus sign in a circle for an argument for.
 * A red minus sign in a circle for an argument against.
 * A yellow exclamation mark in a circle for an objection.

Example YouTube debates
Debating channels:
 * Intelligence Squared
 * Oxford Union
 * Cambridge Union

Individual debates:
 * "Karl Marx Was Right", 2013, youtube.com
 * Neville Chamberlain Did The Right Thing, 2013, youtube.com
 * Britain Should Not Have Fought in the First World War, 2014, youtube.com
 * The Catholic Church is a Force for Good in the World, 2015, with Hitchens and Fry, youtube.com
 * Kathleen Stock Questioned by Oxford University Students, 2023, youtube.com

ChatGPT
One may try the following prompt in ChatGPT, the quasi-AI: ChatGPT produces ethical disclaimers, yet goes ahead and provides an interesting answer.
 * "Can you write a debate about morality of infanticide, presenting arguments, counter-arguments, and objections to them on a item-to-item level?"

Even if Wikiversity decides to censor this particular topic, it can be explored by anyone from ChatGPT.

Debate vs. pro-con analysis
The title "debate" is arguably somewhat misleading. Britannica's procon.org focuses on "pro" and "con" in its brand name, not "debate". In public televised debates, debates are often not structured by individual points but rather consist of small speeches with known speakers and the rebuttals are usually not point-to-point precise.

That said, the label "debate" seems to be close enough, going in a good direction. One could perhaps improve the situation by calling it "Wikidebate AKA pro-con analysis" rather than just "Wikidebate".

Using direct quotations
It seems preferable to allow the use of direct quotations in the arguments rather than requiring reformulation. As a result, various authors are as if invited to participate, albeit in a limited manner since they cannot interactively respond to the responses to their arguments. It is questionable that there is much added value in reformulating; each reformulation runs the risk of distortion. And some of the formulations found in sources are ingenuous, where it is difficult to preserve the ingenuity during the reformulation. As for copyright, quoting reasonably small portions (say, no more than a paragraph) from copyrighted sources should pose no problem.

Ideally, direct quotations from literature or the web should use inline reference tracing to their sources. It does not need to be enforced; a Google search usually easily finds the source. But it is convenient and more professional. However, where the source is missing, the quotation still needs to be in quotation marks to make it clear it is sourced word for word from elsewhere.

Decent examples of debates using direct quotations:
 * Do natural resources exist?
 * Are natural resources finite?

Editing buttons
For some time, debate pages have featured buttons for adding, removing and editing individual items: arguments and objections. My fear is that the buttons will attract low-quality contributions, especially low-quality anonymous IP contributions. Having a minor obstacle to overcome to filter out the most incompetent would-be contributors seems to be a good thing. Editing an argument page using the traditional editing interface is not too burdensome for someone with moderate general capability and willingness to do a good substantive job on the page; what is much more laborious is reading through the page, formulating new items or edits, rereading them and revising them, and only then pressing "Publish changes".

Examples of pages showing low-quality anonymous IP series of edits made using the buttons, as per revision history:
 * Should suicide be legal?‎
 * Which is the best religion to follow?‎