Table of Contents
People focus on special categories of argument. Deduction, induction, abduction, argument from authority, ad hominem argument, non sequitur argument, etc.
But sometimes people are confused by the concept of regular arguments that don’t fit those special categories. Most arguments are just plain arguments, not inductive, deductive or about a fallacy. The standard categories people talk about are incomplete – they cover only a small minority of the arguments people make – but some people believe they are complete or nearly complete.
People have told me that arguments would or could be deductive if they were made more rigorously. A regular argument is merely a deductive argument which takes shortcuts. I’m not convinced. I’ve asked them to write out the complete, deductive version by removing the shortcuts, but they’ve never been able to do that even for one reasonable example. Nor could they tell me of any book with such an example.
The sort of examples I ask for are things that aren’t very math-like. Why is McDonalds a good or bad restaurant? Should I buy an iPhone 13? Should Activision Blizzard’s CEO, Bobby Kotick, lose his job over the toxic workplace culture? Was the ending of Game of Thrones good or bad? I think those are fairly regular topics that people would talk and argue about. A philosophical perspective on arguments ought to address stuff like that, not just stuff like randomly drawing colored marbles from urns.
People make arguments like that I don’t really need an phone upgrade since my current phone is good enough, or iPhones are good (or bad) due to the walled garden aspect. They’ll argue that Daenerys Targaryen’s plot arc didn’t have enough characterization and explanation to get where she ended up, or that obviously Kotick knew what was going on in his company just like Logan on the TV show Succession did. They’ll say McDonalds is gross or unhealthy – or cheap and kids love it.
How is any of that deduction? I think it’s not. Some people tell me it is, but in my experience they cannot connect any issues like those to deduction. Most of them don’t seem to understand how specific or limited deduction is, and actually can’t answer my question: “Please provide a complete specification of the rules of deductive logic as you understand them.” They don’t rigorously know what they mean by “deduction” which is problematic when the point of deduction is rigorous, formal logic which obeys well-defined rules without using intuition. They can’t remove intuition from their arguments but still claim they are using deduction.
Formal logic, by the way, means logic where the argument must be true due to its form, without needing to understand its content. E.g. if you know “If X is true, then Y is true.” and you know “X is true” then you can deduce “Y is true”. That argument doesn’t depend on what X and Y are, beyond them being propositions (statements that can be true or false). This is not the kind of argument people actually make when, e.g., debating the merits of the current U.S. president. Some people claim that’s bad and that regular people are arguing wrong, but I think the philosophers are wrong and should learn to understand regular arguments.
Chinese Food Deduction
A typical sign of a regular argument is the word “because”. A typical argument is:
We should get Chinese food for dinner tonight because it’s cheap and we both like it.
How might you translate something like this into deduction? Here’s an attempt:
- If you are in a particular type of situation, S
- And if people in situation S should get Chinese food
- Then you should get Chinese food
That’s true, given the right kind of situation, but it’s rather vague. What kind of situation is S? Why should people in that situation get Chinese food? It’s a correct deduction but it’s not much use and it’s not very similar to the original argument. Let’s try again by giving more information:
- If you’re hungry
- And if you want to eat
- And if you don’t want to cook your own meal
- And if you care about the price
- And if you want restaurant quality food
- And if Chinese food is cheaply available
- And if Chinese food is nearby
- And if all the people who are planning to eat together are fans of Chinese food and currently in the mood for Chinese food
- And if all the other food options are worse
- Then you should get Chinese food.
We’ve expanded so much on the original argument that it’s hard to tell if its author would even agree with us or would say “That’s not what I meant.”
One of the many problems with this expanded version is that standard deductive syllogisms have two premises. Deduction is all about strict rules, so you can’t just change things around. This doesn’t follow a standard form of deductive logic, so it’s not OK. And, despite all the premises I gave, this is still incomplete. Plus, some of the premises are too vague. Like how close is “nearby”? What degree of hunger counts as hungry? Is this really a universal truth which applies in all situations where the premises are true, with no exceptions? A proper deduction looks like this:
- If all food contains calories
- And if Mongolian Beef is a type of food
- Then Mongolian Beef contains calories
That’s how basic and limited deductions are. Deduction isn’t a freeform thing where you make up arguments while trying to be rigorous. Deduction is some very specific arguments which work due to their form. See the terms “food”, “calories” and “Mongolian Beef”? You could put any other terms instead of them and still get a correct argument.
Here is another deduction with the same form:
- If all widgets contain foozles
- And if a doodad is a type of widget
- Then a doodad contains foozles
The correctness depends on the form of the argument (how things are arranged) rather than the specifics discussed. It still works with foozles instead of calories, and we can see that even without knowing what a foozle is.
The long attempt at a Chinese food deduction isn’t like this. If you replace “nearby” with “foobar”, replace “hungry” with “happy”, and replace “restaurant” with “blacksmith”, you get nonsense. It was the content of the points, not their form, that mattered. It was a list of relevant factors, like hunger and price, rather than a formal argument where only the form matters. So it’s not deduction and isn’t actually even trying to be the same kind of thing that deduction is.
Needing to use judgment – and think about what the words mean and whether the argument makes sense – is a major sign that you’re dealing with regular arguments, not deduction.
So let’s go back to the example argument, which I claim is a regular argument (no special name), and consider what it is:
We should get Chinese food for dinner because it’s cheap and we both like it.
It’s an explanation and it involves implied criticism and an implied challenge.
It’s explaining why we should get Chinese food by giving reasons. The first reason is that it’s cheap. The second reason is that we both like it.
The implied criticisms are that expensive or unliked food are bad. And it’s saying that if only one of us doesn’t like a food, that’s still a problem. Both our opinions matter. It’s ruling out alternatives which don’t fit the reasons given – expensive or unliked food.
The implied challenge is: “Got a better idea?” The argument isn’t trying to be airtight. It’s not the final possible word on the subject. But unless someone has a better idea, it’s adequate to act on.
The argument is aiming to be good enough to take action. It’s implicitly disagreeing with the view that we don’t know enough yet to act and need to do more planning.
The argument relies on the background knowledge of the people involved. They already know a lot about other food options, each other’s food preferences, each other’s budgets, what they ate a lot of recently and got tired of, and more. (Or maybe they don’t know much about that and the speaker is just guessing and hoping this will work. And if he guesses well, the other guy may say “sure”, and if not the other guy may provide new information which changes the analysis.)
The goal of regular arguments is not logical rigor. They’re always pretty incomplete. Although we could be more precise, and go into more detail, that takes effort. We limit our effort to what we think is worthwhile, with the option to expand on what we said if there’s a problem (e.g. the other guy disagrees). This saves us a ton of hassle in the many cases where the other people involved agree with us without further debate.
The goal of regular arguments is to convey some useful information, particularly by explaining reasoning and criticizing alternatives. The information we share is chosen largely for importance and for what we think the other people don’t already know.
Rather than aiming for a bulletproof argument, we normally try to explain enough for other people to fill in the gaps themselves and agree with us. We view them as autonomous individuals of good will, thinking in good faith, trying to figure out a good answer. We are not them and cannot think for them. We can only provide hints and clues to aid in their very complex thought process which we only know a little about. We cannot hand them our thinking. We can only translate a small fraction of our thoughts into English to share, which the other guy then has to translate into his own thinking. The two translation processes are difficult and can easily cause misunderstandings, but we still manage to communicate some.
My buddy is already thinking about what we should eat. He already knows a lot about it. My goal is to share some specific pieces of information that will help him reach a good conclusion. That information can be stuff he hadn’t thought of or didn’t know (e.g. information about my preferences), or information about what areas I’m focusing my thinking on (e.g. he already knows that cheapness is a factor, but by saying it I’m communicating that I view it as a particularly large factor. He can then accept that or else express his disagreement.)
None of this fits the deductive nor inductive models of how we think and argue. Yet this is far more typical and commonplace. It’s important to life, learning, thinking, etc. And actually this view is sufficiently generic that it can subsume the deductive model (and various versions of the inductive model, too, to the extent they are well-defined and logically possible) as a particular special case of this broader model. (As to “abduction”, that is really vague and not very popular anyway.) Deduction is, in short, even when expanded a bit over the original, only stuff that can checked for correctness by simple software – but a lot of what people talk about – and particularly debate about – isn’t like that.
The basic point of arguments is to explain enough useful information for the other guy to understand something for himself. They are like education, like teaching, but sometimes you’re just sharing information he doesn’t have rather than teaching general concepts. You’re trying to help him think, and also you’re trying to be open to letting him help you think, because maybe you’re missing some things that he knows. You can never fill in all the gaps and argue all the details. You have to rely on the integrity of the other guy, and you just help him with the sticking points, the parts where he gets stuck and can’t see how to reach your view, how to deal with some difficulty. But when you convince someone, for most of the stuff, especially the little details, he figures out why you’re right on his own, without being told. You only help him with a few specific cases as well as some general overall guidance. So in a few cases you spell out some details he needed, and also you give an overview which gives him some hints and guidance about all the many cases he figures out himself. You try to give him some useful concepts which he can apply on his own. You try to help him understand a few ways of thinking and arguments and stuff, so he can actually use them on his own initiative, in his own thinking. Deduction is actually really different than this because it only deals with complete arguments.
In general, when you make an argument, it has dual purposes. 1) It could help the other guy think. 2) The other guy could see an error in it, and share that information and thereby help you think. You don’t know, when you say an argument, whether (1) or (2) or both will happen. You don’t know how he’ll respond or who will get the primary benefit. (The primary benefit goes to the person who gets an error corrected. Convincing the other guy to agree with you is a lesser benefit. Fortunately, often, you’re both wrong in some ways, so you can both get primary benefit.)
The Critical Rationalist view is that we learn by argument – by explanation and criticism that, while incomplete, is adequate to make progress, to help people come to agree on more than they did before, to correct some errors, to strive in the future to correct even more errors or reach more agreement, etc. It is these informal (regular) arguments, which care about the specific content brought up more than the form, which are the bulk of human thinking and discussion (and formal arguments are just a special case of them, not technically a separate thing).
By the way, lots of people think “explain” means “predict” or similar, which makes this more confusing. They think of e.g. a mathematical formula as “explaining” any data which the formula outputs. But, in short, explanation is actually about answering “Why?”. It’s about statements with “because”, about reasons, not about math. A math formula, even if correct, doesn’t tell you why that answer is the answer. It doesn’t give you a conceptual understanding, so it’s not an explanation.
Connections with Critical Fallibilism
Critical Fallibilism (CF) claims that all correct arguments are decisive, negative arguments or can be translated to be decisive, negative arguments. These translations are real things that we can actually do today, and there are plenty of examples. They aren’t like the hypothetical idea that maybe, one day, we could somehow translate everything into deduction, even though currently there are zero examples.
So, in short, what is a regular argument? A critical argument. The point of argument is to criticize. Arguments argue against things, either openly or implicitly (positive arguments are negative arguments in disguise, and can be translated if they are actually good arguments).
The point of arguments is to find and point out errors. Argument is part of the process of error correction which is how we make progress. Deduction helps with this because we can criticize ideas which violate the laws of deductive logic.
As a tangent, correct arguments are decisive because there’s fundamentally no such thing as a partial error. When something appears to be partly wrong, you’re bundling multiple things together (some right and some wrong). The world is complex and we often look at lots of complexity at once, so it can contain many errors and many truths, which looks kind of like partial error (the whole group of things is partly right and partly wrong; it partly works and partly doesn’t). If you look at individual things, and clearly differentiate success and failure, then there are no partial errors. However, my claims about decisive arguments are not necessary to my other points in this article.
People often think of an “argument” as anything someone would say during a debate (a.k.a. an argument). But I think that’s inaccurate. Besides (critical/negative) arguments, there are (positive) explanations. An explanation says how or why something works. Although I didn’t make this distinction above, I don’t think explanations of how things work should be called “arguments” because they’re saying what your plan/solution/idea is – sharing its content – rather talking about it. Arguments are meta ideas which argue about whether some other idea is bad (or argue about how good it is, if you accept partial and positive arguments). Whereas saying what an idea is and how it works is sharing the idea itself, not arguing about the idea.
Suppose my goal is to eat. My idea is to open a can of soup, put it in a bowl, microwave it, then eat it. I don’t consider that statement an argument. It’s the idea itself. It’s the thing that can be argued about. It’s just an explanation of what my idea is (a pretty short one that leaves a lot of the “how” implied since it’s common knowledge in our society).
Suppose my goal is to make a robot vacuum, like Roomba, but before that existed. Potential business partners ask me questions about it. How will it work? It will have a battery, motor and wheels so it can move around by itself, I explain. It will have software that controls where it goes. In those replies, I’m explaining the idea itself – what is the idea? how does it work? – not making arguments.
People often mix explaining (talking about an idea itself) and arguing (about another idea) together, including sometimes doing both in the same sentence. I think that has blurred the concepts together in many people’s minds. I blurred them earlier in this article because it was convenient when I was talking about other issues. Also, arguments are ideas themselves and we need to explain/share what they say, why, how they work, etc. It’s sometimes unnecessary to clearly distinguish between the content of ideas themselves and arguments about them, but it’s good to know how to. It can help sometimes.
In conclusion, most arguments people use are not deduction or approximations of deduction, and that’s OK (nor are they inductive nor abductive arguments, which Karl Popper and I have criticized elsewhere). I think regular arguments are critical arguments which talk about why ideas fail at goals. Arguments argue about other ideas, which is different from explaining the content of an idea itself – what it is, and how and why it works – which is not argument. Regular or critical arguments aren’t inferior to deduction – just different – and we shouldn’t look at deduction as the ideal. The common sense concept of an argument – which people use all the time – is not deductive and it’s basically fine and reasonable.