19

I have trouble with many so-called 'fallacies', and the 'No True Scotsman' fallacy is no exception.

Let me quote a famous line from Game of Thrones:

Any man who must say "I am the King" is no true King.

Did Lord Tywin commit the 'No True Scotsman' fallacy here?

Or another less famous quote:

Knights are sworn to defend the weak, protect women, and fight for the right, but none of them did a thing. Only Ser Dontos had tried to help, and he was no longer a knight. No more than the Imp was, nor the Hound … the Hound hated knights … I hate them too. They are no true knights, not one of them.

Did Sansa commit the No True Scotsman fallacy?

All this seems extremely counter-intuitive to me. Both Tywin's reasoning and Sansa's reasoning seem sound to me.

It seems to me that the pattern of speech condemned by the proponents of the 'No True Scotsman' fallacy goes like this:

X is Y, but X is no true Y.

This seems fallacious, as it tries to draw some middle ground between is and is not. I can understand the mathematical reasoning that X is Y if and only if X fulfills the definition of Y; anything more just fuzzies the issue and is fallacious.

But what if we treat 'X is Y, but X is no true Y' as a shorthand for a longer, though more precise version of the same statement? 'Y is a word with two definitions: Y₁ and Y₂. Thought X fulfills Y₁, X does not fulfill Y₂. And in the context of our discussion I consider Y₂ to be more relevant than Y₁.' This is valid, as words do indeed have many definitions.

In this light both Tywin's and Sansa's statements immediately seem sound. Specifically, both Twyin and Sansa seem to be pointing at the distinction between de iure and de facto.

A person is de jure a King if he's crowned a King. However, the office of a King was established to formalize supreme authority. Therefore, if a person is de jure a King, even though he has no actual authority (because he's just a figurehead), then he's not de facto a King. We may use a verbal shorthand and say he's not a true King.

Likewise, a person is, at least in theory knighted so that he can defend the weak, protect women, and fight for what is right. If a person is de jure knighted, but does not defend the weak, protect women nor fight for what is right then we can say that he's not de facto a knight. Again, we may use a verbal shorthand and say he's not a true knight.

Is this correct? Or did I make a mistake and is No True Scotsman indeed a fallacy?

7
  • 3
    In both your examples "true Y" stands for something different than "Y", in which case it's not a fallacy, but a semantic argument.
    – rus9384
    Commented Jan 27 at 23:36
  • 34
    The No true Scotsman pattern is not "X is Y, but X is no true Y", but "If X is A then X is no true Y", where A is contrived to suit the speaker's purposes. In other words, Y is redefined into "true Y" by making A the litmus test for it. Statements of the same form may well be valid when A is part of the conventional definition of Y. In that case, they simply register that X fails the definition of Y, and "true" is added merely for emphasis. And this is a general feature of informal fallacies, they cannot be distinguished from valid statements by their form alone.
    – Conifold
    Commented Jan 28 at 2:34
  • 9
    None of these contexts are logical arguments between two or more people where some kind of debate happens. These are not fallacious because the very basis for it possibly being a fllacy is missing.
    – Polygnome
    Commented Jan 28 at 11:09
  • 8
    I can't see how your example from Game of Thrones should be an instance of the No True Scotsman fallacy. The mere presence of the words "no true ___" isn't what makes the fallacy.
    – kaya3
    Commented Jan 28 at 11:32
  • 3
    I believe the crux of the fallacy is contriving classifications to not need to discuss something without actually addressing the reclassification or matter at heart. Though the line between substantial and contrived is ultimately subjective things can be done in good faith or bad faith. In that sense, your true knight quote does not fit this because it's pretty clear that what is being said is basically "actions make a knight a knight, and not the title".
    – DKNguyen
    Commented Jan 28 at 18:58

10 Answers 10

74

To my understanding, the no true Scotsman fallacy is not captured in the syllogism

  1. No Scotsman microwaves his tea.
  2. Angus microwaves his tea.
  3. therefore Angus is not a Scotsman.

Rather, it may be found in the following dialog

  1. Ewan: No Scotsman microwaves his tea.
  2. Richard: Angus microwaves his tea.
  3. Ewan: Angus is not a true Scotsman

To my understanding, the no true Scotsman fallacy is an example of shifting the goalpost.

Ewan has made a universal claim about Scotsmen. Richard has provided a counter-example. Ewan, instead of acknowledging that his claim was over-broad, and that there are exceptions, doubles down and denies that Angus is a "true" Scotsman. A "true" Scotsman is sufficiently ill-defined that Ewan can maintain its truth in the face of a counter-example of a "supposed" Scotsman who microwaves his tea.

The statement

A true knight helps the weak.

to my mind is NOT an example of the no true Scotsman fallacy (or any fallacy for that matter). It is simply an expression of the belief that a knight who fails to help the weak has failed to do what the utterer believes is a knight's duty. The utterer of the above statement is giving a partial definition for the term "true knight", not weaseling out of an offered counter-example.

6
  • 23
    +1. And worth adding, the strong superficial resemblance between the given (valid) syllogism and the dialogue illustrating the fallacy is no coincidence: the famous named fallacies are not arbitrarily-chosen invalid arguments, they got named exactly because they’re superficially convincing forms of argument which are slightly subtle to rebut, and hence need a special mechanism for clearly identitfying them when they’re used in debate. Commented Jan 28 at 18:44
  • So the "no true scotsman" version of the knight exchange would be: "Knights always protect the weak -> In this incident they didn't do anything to prevent weak people from getting slaughtered -> The knights involved in this incident weren't true knights".
    – Philipp
    Commented Jan 29 at 15:11
  • 7
    @Philipp I don't think it is merely the form of exchange that makes a no-true-Scotsman. I think intent to cover a mistakenly made over-generalization is essential. Consider this exchange: "Damsel-in-distress: Why did you rescue me, Sir Kay? Sir Kay: Knights always protect the weak. Damsel-in-distress: Brian de Bois-Guilbert did not protect the weak. Sir Kay: Brian de Bois-Guilbert was no true knight." The form of the last 3 statements fits the pattern. But I don't think Sir Kay was (necessarily) attempting to cover a mistake. His initial statement was a forgivably imperfect explanation. Commented Jan 29 at 15:37
  • 1
    No philosopher would dare use the "No true Scotsman" fallacy to justify an argument. Richard: "Ewan is a philosopher, and he did while arguing about tea." - well... no true philosopher would, anyway... Commented Jan 29 at 19:26
  • It occurs to me to wonder about the case where Ewan is the interior minister in a future independent Scottish government and has the power to grant or rescind Scottish citizenship; or where Sir Kay has delegated authority from King Arthur to award or rescind knighthoods. Commented Jan 30 at 10:36
27

Others have said that it's a case of moving the goal posts, which it certainly is. If nothing else, that makes it poor sportsmanship in a debate format. Still, it does happen from time to time that the goalposts were in the wrong place.

  • All scientists agree that smoking causes cancer.
  • What about Astrophysics Andy?
  • Ah, sorry. Andy wisely says he hasn't studied cancer. I should say all medical scientists.
  • What about Paid-off Paul?
  • OK, so all medical scientists surveyed not in the employ of tobacco companies.

The goalposts have been moved, but this seems on the face of it to be a perfectly reasonable response if you're truth-seeking and not debating.

Now, suppose instead the response were

  • OK, so all true scientists.

The alarm bells sound. Specifically, to me this feels like I'm about to be presented with either circular reasoning or equivocation. The modifier "true" typically means in practice "the ones who fit my pattern". So, it will be circular reasoning if you update all your premises to use "true", and equivocation if you don't. That is, if you're willing to say "All true scientists, that is scientists who think smoking causes cancer, think smoking causes cancer." then you'll get called out for the obvious circle. If they're really on it, they'll ask why we should care for the opinion of some arbitrary and possibly empty set of scientists you call "true". Now if instead you try to slip through some semantic sleight of hand that you used "True scientists" in one premise (the one where you try to get the evidence to be true) and just "Scientists" in another (the one where you try to get the evidence to matter) it's going to be equivocation.

All that is to say, you're absolutely right that at the heart of a no true Scotsman argument sits two different definitions, and typically something of an attempt to avoid meaningfully distinguishing them. Perhaps the best question in response is to ask "So supposing I should encounter Scotsman, how would I know him to be true or false?" If the answer is "He wears a kilt" and the matter under dispute is something like "Kilts should be Scottish school uniform" then you probably have a circularity problem. On the other hand if the distinction given is "He's chucking large logs around for sport" then you can proceed with the discussion, although with some caution around whether caber tossing is a relevant distinction to draw, and whether anything is subtly imposed on the perhaps more bookish "false" Scots which only actually holds for their tree throwing cousins.

1
  • 5
    +1. It basically comes down to re-defining or re-classifying to avoid addressing some crucial topic at hand which I can't imagine being done in anything other than bad faith.
    – DKNguyen
    Commented Jan 28 at 18:59
12

"No True Scotsman" is usually an informal fallacy. An informal fallacy is a mistake in argument when the form of the argument is valid, but there is a problem in content or context, such as: a false premise; a misleading definition; a misuse of language such as equivocation of two different meanings of the same word; or a wording which implies a conclusion other than what the argument actually shows.

By contrast, a formal fallacy is an identifiable error in the form itself.

Consider:

1. No Scotsman microwaves his tea. 
2. Angus microwaves his tea. 
3. therefore Angus is not a Scotsman. 

It is a perfectly valid argument - as long as we agree that to be a Scotsman is to not microwave one's tea, we agree that Angus is not a Scotsman, even if Angus was raised in Scotland by people raised in Scotland by more of the same back many generations. But in practice, this employs a misleading definition. Normally people mean men like Angus when they refer to Scotsmen, so our nonstandard definition is misleading. As a result, our wording implies a lurid informal conclusion like "Angus must be lying about having been raised in Scotland," which goes far beyond the boring, definition-based conclusion that the formal argument actually shows.

On the other hand:

1. No Scotsman microwaves his tea. 
2. Angus is a Scotsman. 
3. Angus microwaves his tea. 
4. 2 with 3 implies a Scotsman microwaves his tea. 
5. 1 and 4 are a contradiction. 
6. Therefore 5 implies 2 is false. 

It has a formal fallacy: 5 implies 1 and/or 2 and/or 3 is false, not just 2. But the people discussing Angus' tea-brewing habits are unlikely to speak in such a way that a formal fallacy can be identified.

4
  • No Scotsman microwaves his tea. Angus microwaves his tea. Therefore Angus is not a Scotsman. - the reasoning here seems pretty clear to me: By custom Scottish people do not microvawe tea, therefore Angus does not follow Scottish customs. And indeed for some poeple following national customs is an important part of national identity, sometimes even more than the 'correct' birth.
    – gaazkam
    Commented Jan 27 at 20:19
  • 2
    @gaazkam But that would bring about questions as to who defines what these customs are and aren't and who is in the group of people to define that. Either way there is an (unstated) use of an uncommon definition.
    – haxor789
    Commented Jan 27 at 22:09
  • My father was a Scotsman, to be sure. Me mum a Scotswoman. Commented Feb 12 at 4:13
  • ... cont. sorry for break line... My Pops, true to form, would never think to microwave a tea. He would however, without noticing, drink tea microwaved by me Mum. Now if this "dont brew tea in a microwave" stands, where does the fate of my father lie? Is he kicked out of the clan, for actions unaware? Or is the distinction between microwaving, and consuming microwaved product enough to buy him a bye? Inquiring minds need to know. Commented Feb 12 at 4:19
3

Formal fallacies are pretty straight forward, you just need to check whether the conclusion does necessarily follow from the premises and if it does it is a valid argument. If it furthermore happens that the premises are also true (and thus the conclusion is also true), then the argument is called sound.

This however is a so-called "informal fallacy" which is a quirk of natural language which lacks the unambiguous clarity you'd find in pure logic.

The classical example of an informal fallacy is probably the fallacy of equivocation e.g.:

Since only man is rational.

And no woman is a man.

Therefore, no woman is rational.

Which is valid, if only men were rational (excluding everything non men) and women are "non men", then women are not rational. In this case the conclusion follows from the premises so the argument is valid.

Now obviously the conclusion is not true so at least one of the premises must be false (because if they were true the conclusion would also NEED to be true).

Now the rhetorical trick of the fallacy of equivocation is that the premises are evaluated one by one in their own particular context.

So the first premise is evaluated with reading "man" as "human" and so "only the human species is rational", which for the sake of argument we read as true. While the second premise is evaluated reading "man" as "male", which for the sake of argument we also read as true. So valid statement + true premises = true conclusion so a sound argument?

The problem is that they are only true within their context of interpreting the word "man" if you used male for both cases then the first premise would turn out to be false and also directly contradicting the conclusion, thus the argument would be valid but not sound. If you used human in both premises then the first would be true but the second one would be false thus again valid but not sound. And if you keep them as "human" for the first premise and "male" for the second, then both premises would be true, but the argument would no longer be valid, thus also not sound and not even valid.

And a no true scotsman fallacy is a similar slight of hand.

You say no X does Y

Are presented with an example of X doing Y

and thus ad-hoc rectify your claim with but they are no TRUE X

Which again uses the equivocation of "X" and "true X", which are equivalent or easy to confuse unless you provide a clear definition for each. So if you were to give a definition of "true X" and plug that into the whole thing then you would be correct:

So if Sansa says "(True) Knights are sworn to defend the weak, protect women, and fight for the right" and you list a bunch of knights who don't then Sansa is correct to dismiss your knights as no true knights, because they fail to adhere to the previously established definition. They aren't knights in the sense of the definition that she gave you of what a knight is.

The problem is that this is usually just the first half of that argument, the latter half being "... And because we've established them not to be knights they should not get this particular privilege of knights." for example.

And then you're back to a fallacy of equivocation because Sansa's definition of a knight might be different from the one that is used to justify this privilege.

Or the other problem. A knight is accused of a crime and the response is "knights don't do crime", followed by a proof that they did indeed commit the crime upon which you're met with the remark of "then they are no true knights".

Where again something somewhere went wrong with that argument but what in particular actually depends on the specific context. Like is knight to defined as a property of character? Is it a certain status? Is the mutual exclusivity knight and crime one of virtue, of law of impossibility? Does the crime invalidate the knighthood or is the knighthood status persistent even after having committed a crime? And most importantly does your definition of knight stay consistent throughout the argument.

So yeah these informal fallacies are kinda tricky because there can be a fallacy, there doesn't have to be one (your example of Sansa), there could be deliberate or accidental obfuscation by lacking,improper and/orambiguous definitions. There can be cases of changing definitions. Like a knight could refer to a much bigger group then a "true" knight and that can have implications further down the line of the argument or it could even be correct to dismiss counter examples as "no true X" if they indeed aren't. So that really depends on the specifics.

2
  • 2
    @ScottRowe If I'm reading it right, he's not saying it's inappropriate to do so in those circumstances, however it's only appropriate because there is a common definition that sets the standard for membership as defined in that context. Everywhere else needs to be taken on a case-by-case basis: Do we have a common understanding for the context and definition? If so, it's fine; otherwise it's potentially deceptive and falls under the definition and rationale of this fallacy.
    – Kaji
    Commented Jan 28 at 6:17
  • 1
    @ScottRowe I think the biggest problem with this "fallacy" has nothing to do with logic itself but is more concerned with the question of "Who has the authority of definition (about who is and isn't X or a member of X)" and as that is something that touches upon the question of identity and self-identification or the lack thereof and as such it is almost inherently controversial. Usually people define themselves as a) member of a group and b) as virtuous. Which they use as synonymous as long as both of them are true, but who gets to decide whether they should leave a), b) or both if it doesn't
    – haxor789
    Commented Jan 28 at 12:08
3

Let's look at your example:

'Y is a word with two definitions: Y₁ and Y₂. Though X fulfills Y₁, X does not fulfill Y₂. And in the context of our discussion I consider Y₂ to be more relevant than Y₁.

This is completely fine, but it's not the No True Scotsman fallacy. Suppose instead that you had just asserted that everything meeting definition Y₁ also satisfied (as an empirical fact, and not by definition) one of the conditions of Y₂. And suppose I find a counterexample. Then it's wrong to claim you meant definition Y₂ all along. After all, if you did mean definition Y₂ all along, then your initial claim was not interesting.

Arguably, this is at least related to the Motte and Bailey fallacy/doctrine. Roughly speaking,

Y₂ implies Y₂

is the very boring motte, and

Y₁ implies Y₂

is the bailey.

3

What makes it a fallacy?

The fallacy is one of equivocation, usually conflating "true X" and "X". This then conceals a false premise, usually either "my approval determines who can claim to be X", or "doing Y determines who can claim to be X".

If the "true X" and "X" are conflated, and the hidden premise is false, then it is an instance of the fallacy. Otherwise, it is not.

  • Given: No true Scotsman microwaves his tea.
  • Given: Angus microwaves his tea.
  • Then: Angus is no true Scotsman.

Filling this out without equivocation, to show the meanings of "no true" and to fill in the missing premise:

  • Given: No Scotsman I approve of microwaves his tea.
  • Given: Angus microwaves his tea.
  • Given: Being Scottish is dependent on one of:
    • My approval
    • Tea preparation
  • Then: Angus is not Scottish.

The falseness of the implied premise then becomes clearly visible: neither of the options determines who's Scottish.

Let us look at the quotes you give.

Lord Tywin

  • Given: Any man who must say "I am the King" is no true King.
  • Given: Joffrey says that.
  • Then: Joffrey is not king.

This could be rephrased in a couple of ways.

  • Given: No King I approve of says "I am the king"
  • Given: Joffrey says that.
  • Given: Being king is dependent/defined by on one of:
    • My approval
    • Them not mentioning that they're king
  • Then: Joffrey is not king.

Has "not acknowledging their position" been a necessary requirement to retaining one's crown, in any country, in any period of time, ever? No: many good and true kings have had to argue and fight for their position.

So, if that had been what he had meant, then yes: Lord Tywin would have committed the 'No True Scotsman' fallacy.

But if I remember right, in context he meant something more like:

  • Given: Appealing to the crown's authority for obedience is a crutch for weak kings.
  • Given: Joffrey demands obedience just because he's king, and instills no loyalty.
  • Then: Joffrey will make a very poor (and short-lived!) king unless he works on that.

In that context, there's no implication that Joffrey isn't king, or that acknowledging his crown would or should disqualify him from wearing it. Here, "true king" isn't conflated with "king".

Instead, it's an exhortation to avoid over-reliance on the authority of the crown, but that advice came dressed in the language of a no-true-Scotsman, making it easy advice to ignore.

Given this interpretation, then no: Lord Tywin would not have committed the 'No True Scotsman' fallacy.

Sansa

Knights are sworn to defend the weak, protect women, and fight for the right, but none of them did a thing. Only Ser Dontos had tried to help, and he was no longer a knight. No more than the Imp was, nor the Hound … the Hound hated knights … I hate them too. They are no true knights, not one of them.

This is a slightly different formulation. Breaking that down:

  • Given: Knights are sworn to {protect}.
  • Given: None of them {protected}.
  • Then: They are not knights.

Again, that can be interpreted a couple of ways. One is still equivocation, concealing a hidden premise:

  • Given: Knights are sworn to {protect}.
  • Given: None of them {protected}.
  • Given: Being a knight is dependent on upholding their vows.
  • Then: They are not knights.

How true that premise is, depends on the law of the land. There are many occupations which require swearing an oath on entry, but the penalty for breaking one's vow varies from occupation to occupation. Typically, removal from the group only happens after investigation is complete, rather than being retroactive to the time the vow was broken. So, even if they might in future be investigated and kicked out of their order for such oathbreaking, they were still knights at the time he spoke.

If this had been what she'd meant, yes, Sansa would have committed the no true Scotsman fallacy.

But again, I think she meant "no true knights" to mean "no knights who're true to their word", or "faithful knights". She wasn't conflating "true knights" with "knights", but rather was basically saying "ACAB".

This is a slightly different formulation. Breaking that down:

  • Given: Knights swore an oath.
  • Given: None of them present upheld that oath.
  • Given: It's reasonable to extrapolate the behavior of the knights present to all knights.
  • Then: There are no knights who are true to their word.

I think that's closer to his meaning, and in that case, no, Sansa did not commit the no true Scotsman fallacy.

TL;DR: Conclusion

  • No true no-true-Scotsmanner avoids conflating "true X" with "X".
  • Tyrion and Sansa were not conflating those terms.
  • Tyrion and Sansa are no true no-true-Scotsmanners.
1
2

You're basically right, but what distinguishes a No True Scotsman fallacy is that the property in question would previously not have been considered as part of definition.

As you said, in both of your examples, it is plausible to claim that the properties in question really are essential for what it means to be king / knight respectively, in that sense these are no Scotsman-fallacies.

  • A king, or generally a leader of state, does not rule because he has god-given strength to force his subject to carry out orders. He rules because most of his subjects agree that he's the king and consequently follow out his orders and give him power this way. In short, the definition of "king" is indeed that people already know he is the king, from which is it valid to infer he doesn't need to explicitly say "I am the King".
  • For knights, here we're basically talking about chivalry. If knights are sworn to defend the weak etc., then this again becomes part of the definition for what is and what isn't a knight, and thus is fair to put in question whether someone not adhering to the commandments of chivalry still qualifies to be a knight.

These are well-established principles.

In a No True Scotsman, it's different: the definition of who is a Scotsman is straightforward, it's just somebody from Scotland. Crucially, when Ewan makes the statement "no Scotsman microwaves his tea", he refers to this standard definition, i.e. what he asserts is "no man from Scotland microwaves his tea". This is presented as a tautology (e.g. microwaves are forbidden in Scotland -> when you're in Scotland you can't microwave your tea) or something he has observed ("I have asked every Scotsman whether they microwave their tea, and all answered no").
Then, where the fallacy is committed is that upon being confronted with contradicting evidence, Ewan ad hoc makes a change to the logic by promoting "does not microwave tea" to part of the definition of a Scotsman, in other words turning a proposition into an axiom.

1

The bare statement "No true knight" is not a fallacy. It would be a fallacy if used in defense of a general claim about knights. The sentence "No true knight would ignore the plight of the innocent" is just the expression of an opinion about how a knight should behave, but if the conversation were to start with Sansa saying, "No knight would ignore the plight of the innocent." and then someone gave a counter-example where a knight did ignore the plight of the innocent, if Sansa were to respond, "Well, he was no true knight", then Sansa would be committing the fallacy. Sansa made a general claim, a counter-example was proposed showing that the general claim is false, so Sansa tried to modify the general claim without admitting that the original claim was in error.

3
  • 1
    So are there any general claims which can stand, and be used in determining things? If we just stick to those claims, things would go better.
    – Scott Rowe
    Commented Jan 27 at 22:22
  • The problem is, it's shorthand for "Nobody I would consider a true knight", glossing over the question of what that definition is and whether anyone else would agree with all its details. Don't confuse human speech with formal logic; they have very little to do with each other, alas. (And philosophers have had a bad habit of making exactly that mistake, analyzing a subset definition and thinking they've proven something about the more general term.)
    – keshlam
    Commented Jan 28 at 5:14
  • @keshlam, whatever it is, the statement used alone rather than defending an unrestricted universal, is not a fallacy. Commented Jan 28 at 5:51
1

I think the difference is in the GOT quote:

Any man who must say "I am the King" is no true King.

The speaker is talking about the attributes of the man, denigrating him and saying his actions are unbefitting those of a King.

The subtle difference I think in statements deserve to be designated the 'no true Scotsman' fallacy are when you are declaring attributes of the category and ignoring counter-examples:

All kings are just.
What about King Joffrey?
Well, he was no true king. My statement that kings are just still stands, we just won't count Joffrey as a true king in spirit.

This line of reasoning is arguing that kings are just by nature, and ignoring a counter-example. The GOT quote is arguing that King Joffrey does not act in a kingly manner, but not declaring that is is not technically a king. Rewording the argument:

Any man who must say "I am the King" is no true king.
What about King Joffrey? He says that.
My point exactly, I am saying he is not a good example of a king.

0

No philosopher thinks the True Scotsman Fallacy isn't one.

But gaazkam does!?

Well, gaazkam is not a true philosopher ;-).

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .