|Cogito ergo sum|
Logic and rhetoric
|—Ray Comfort, on the half-muppet brainchild he farted into existence with the help of co-genius Kirk Cameron|
A straw man or Aunt Sally (mostly UK) is a logical fallacy which occurs when a debater misrepresents their opponent's argument as a weaker version, sometimes intentionally, and rebuts said version — rather than their opponent's genuine argument. Intentional strawmanning is usually done with a certain goal in mind, including:
- Avoiding real debate against an opponent's real argument, because the misrepresenter risks losing in fair debate
- Making the opponent's position appear ridiculous as a way of poisoning the well
Unintentional misrepresentations are also possible, but in this case, the misrepresenter would only be guilty of simple ignorance. While their argument would still be fallacious, they can be at least excused of malice.
The "reflexive" straw man
The straw man appears to be one of the most frequently mentioned fallacies on the contemporary Internet. Sometimes, supposed examples can themselves be straw men, i.e. if one claims something is a deliberate straw man, and it isn't, then you yourself are creating a straw man.
- straw dog
- wooden dummy
- Aunt Sally
- hollow man
- man of straw
Most of these names rely on the notion that the argument proposed can be easily knocked down. An "Aunt Sally" is originally a hollowed out coconut at a British fair, often with a face painted on it, which is set up in order to be knocked or shot down for a prize.
A straw man debate takes the following form.
- Person X raises argument A1.
- Person Y describes argument A1 as argument A2, which appears similar to (but is substantially different from) argument A1.
- Person Y refutes argument A2.
- Person Y claims that argument A1 has been refuted.
In a logical syllogism, this takes the form:
- P1: X (their argument) is Y (the strawmanned argument).
- P2: Y (the strawmanned argument) is false.
- C: X (their argument) is false.
The syllogism is valid unless X is not Y; in a strawman argument, X and Y are different.
Consider that the range of statement forms that could be made about something:
- All X are Y.
- There are X which are (not) Y.
- Few X are (not) Y.
- Some X are (not) Y.
- Most X are (not) Y.
- No X are Y.
It is common to represent an opponent's position as being an extreme position, e.g., that pro-choice advocates think all abortions are acceptable, or that anti-abortion advocates think no abortions are acceptable.
This is because it is easiest to defeat the forms near the extremes and hardest to defeat the forms near the middle. For example, to defeat "All X are Y", proof of only one counter-example is needed (namely, one statement of the form Some X are not Y). Conversely, to defeat Some X are Y, proof of a complete lack of examples is needed (namely, that "No X are Y"). Proving categorical statements is harder than taking them down.
Steelmanning is a neologism for the principle of charity, coined on LessWrong by Scott Alexander, in analogy to "straw man". The principle of charity is the maxim that in debate, one should take the strongest possible understanding of the opponent's position. It is a warning against the straw man argument. Only if you defeat the strongest possible version of your opponent's argument can you say that you've robustly defeated their position.
Related to steelmanning is the Ideological Turing Test, which asserts: If someone understands their opponent's belief so well that they can convince a blinded observer who is a true believer, and yet still not actually believe it, there's a good chance that that belief is wrong. The first part of this test is useful: if you can repeat somebody's argument back to them in such a way that they agree with everything you say (and do not wish you had included more), then you have properly understood/summarized their position. However, the second part is not: just because you can understand somebody else's view fully, but still reject it, does not mean that their belief is incorrect. This is because your rejection of their belief is based on additional beliefs that you hold, which may be incorrect or which they may not accept.
See also Dennett's Rules, which includes a version of steelmanning.
Unlike steelmanning, Ironmanning (which is in no way related to the superhero Iron man) is the converse of strawmanning. An ironman argument is an intentional misrepresentation of one's own or another's position, often used to present one's own or another's stance as being far more defensible or uncontroversial than it actually is, often to make it appear that one's critics are unfair, shrill, or even outright evil.
The title of the argument comes from the art of practising fighting techniques against men made of straw: which is a problem in that straw men don't fight back, don't wear armor, don't bleed and generally aren't anything like the sort of thing you would actually encounter in a battle. Therefore someone arguing against a straw man is just arguing against an idealised opponent that only exists in that person's own head.
Straw men are notoriously easy to construct, and require little more than extending the opponent's arguments beyond their original point until their stance appears ridiculous — appearing like the fallacious use of reductio ad absurdum. Once the opponent has accepted (or failed to refute) such a set-up, one can simply attack the straw man position instead of the opponent's actual points, and claim any subsequent attempt to correct the situation as conceding the argument.
Sometimes it's so easy, you can do it without thinking.
- 12 Arguments Evolutionists Should Avoid
- How come there are still monkeys?
- Peanut butter argument
- Random chance
- "Atheists hate God."
- "Atheists believe in nothing."
- "Atheists reject Christianity specifically: why aren't you getting mad at Muslims?"
There are enough strawman versions of atheism going around to fill a whole article itself. This is most likely due to the fact that religious apologists who generate these arguments have never been atheists (or possibly exaggerate their 'atheism' if they have been), in contrast to many atheists who were previously believers.
What usually unites straw arguments against atheism is that they're constructed when a specific religion thinks that it's all about them — thus ignoring the main point of atheism, that it's the rejection of all religious belief equally. Further common ones include the assertion that atheists "believe in nothing", which seems to confuse it with the distinct concept of nihilism (even then, "believe in nothing" is a slight strawman of nihilism) or that they lack a moral code or any concept of morality. The latter either stems from or forms the base of the argument from morality, but is easily refuted simply by looking at the behavior of atheists, which is most conveniently done by examining the crime statistics of secular countries.
Other examples are:
- Atheophobia — reasons given for hating atheists
- Misconceptions about atheists
- Misotheism — hatred of god(s)
Straw man arguments against religion are also easy to generate — conflating religions together into one entity is perhaps the most common. Zen and Christianity have almost nothing in common, for example, yet both can be dismissed readily by a straw man of a generic "religion" concept.
Defining anyone who is even moderately religious as an unthinking, uneducated sheep-like entity who blindly accepts whatever their priest or pastor tells them is another. In this case, the straw man is to conflate fundamentalism with mainstream belief — decent sites such as Fundies Say The Darndest Things on occasion have an unfortunate tendency to mistake anyone talking about prayer, faith, or Jesus as rabid fundamentalists. Straw man positions on religion are often held by anti-theists (who generally are, let's face it, teenagers who have just discovered the internet).
Prominent biologist Richard Dawkins has been accused of generating straw man religions to fight in his pro-atheism writings such as The God Delusion by falling into a few of the traps discussed above.
For a good example of lumping religion into one, type "science vs religion" into Google Images. What you will find is a large collection of memes that compare science with fundamentalist, Young Earth creationist Christianity. None of them take into account that most Christians don't accept the Old Testament or Genesis creation myth as literal fact, nor the fact that religion is not an alternative to science, since it is possible for a scientist to believe in God. The only other religion mentioned is Islam, along with the fallacy that "religion flies you into buildings", omitting the numerous different things religions believe about the Earth's origin, age etc. "One size fits all" will not do in criticizing religion, generally speaking.
In the political sphere, the easiest way to construct a straw man is through the generous application of stereotypes. By making the assumption that a politician believes in all the ideologies associated with their general political leaning, they can be dismissed more easily. For instance, assuming a fiscal conservative is also socially conservative or assuming a pro-choice politician is also in favour of extreme wealth redistribution are common fallacious straw men that are made by politicians, pundits and voters themselves. For the left, a common straw man is assuming that anti-abortion advocates must be "pro-life" on all issues (such as opposing capital punishment), not just abortion, and calling them hypocrites if they are not (although some anti-abortionists advocates do oppose both). This is no more true than the example of pro-abortion advocates having to favor choice on all issues, not just abortion. The same is true of conflating ideologies, such as stating that liberalism, socialism and communism (two of which are economic systems, the other being a form of thought in support of capitalism) are identical (a rhetoric notoriously popular in the US).
This is extremely problematic in politics because it is not possible to give an accurate appraisal of political beliefs without knowing what those beliefs are, or by presuming that belief A automatically means subscribing to belief B. This is also used within politics to introduce snarl words into discussions.
It is quite easy to make a straw man argument of woo explanations and things like alternative medicine. The classic example is homeopathy, an alternative medicine treatment where active ingredients are diluted substantially, on the principle that the more dilute it is, the more potent it is. In fact homeopathy often takes this to the point of diluting the substance until no trace of it remains in the remedy. It is, of course, absurd to think that this could work, but it is also a straw man argument because homeopathy includes a succussion procedure involving striking the solution in a special way. Thus, homeopathic remedies aren't "just dilutions" but dilutions that have been whisked about a bit. Woooooooo!
It is contentious whether this is a true straw man position, as the burden of proof lies with homeopaths to establish that the succussion procedure adds anything new to the remedy, such as instilling "water memory". Thus, given the balance of evidence, homeopathic remedies are just dilutions. Skepticism, however, requires an honest appraisal of all the facts and ignoring the striking involved in succussion, even if it is just an escape hatch, is still wrong in principle. Remember extraordinary claims require extraordinary evidence.
Particularly on the internet, it is easy to create a straw man argument about the feminist movement by jumping to generally inaccurate conclusions and cherry picking (particularly involving extreme views held by a minority). Such arguments include platitudes like "feminists want women to rule over men" or "feminists are 'reverse sexists'". This almost inevitably involves either the invocation of the microblogging platform Tumblr or the invocation of Andrea Dworkin (to which the misattributed quote "all heterosexual sex is rape" goes), but not often the invocation of her genuine positions.
Not quite strawmanning
There are a few practices that are sufficiently close to being strawmanning that they can easily veer into the full strawman fallacy, but can also completely avoid it. As a a general rule, the clearer you are in your mind and argument that you are doing one of the following and are only discussing a hypothetical or naive position, the less likely you are to fall into the fallacy.
- Discussing a hypothetical position you understand your opponents or listeners do not hold, in order to get at a definition of positions they do. (It's preferable to avoid this, but it can be necessary when, for example, dealing with somebody whose position is vague or unknown even to them.)
- Attempting to take apart a methodology or a philosophical position, in order to expose hidden connections (e.g., "if you take Assumption A without Limitation B, you get Undesirable Result C").
- Discussing a position that, while horrible and (hopefully) not held by your opponents or listeners, is still a logical objection to your own (e.g., "If you take the assumption that 'murder is perfectly okay', I'll grant you that my argument against the death penalty completely falls apart… although if you hold that as a general assumption, I'd rather not meet you in person, for obvious reasons.")
- Discussing an explicitly strawman version of an argument, either while describing a third party's position ("A's description of B's argument is a strawman, and here's why"), or as part of a discussion of whether or not a counter-argument to the actual position is valid ("While this argument may work against the 'naive' version of the position, let us see how it fares against B's actual argument").
Further, discussion of fringe positions that are actually held, but only by people of dubious sanity, can lead into the straw man fallacy, through at least two vectors:
- Implying the position is held by a larger or wider group than it actually is.
- Reading too much into a position held by a crazy person. (See this section for a good example of this phenomena.)
- Ad hominem
- Gish Gallop
- Logical fallacy
- Psychological projection
- Suppressed correlative
- Term of art
Want to read this in another language?
- Strawman, Fallacy Files
- Your logical fallacy is strawman, Your Logical Fallacy Is
- Straw Man Arguments: How to Recognize, How to Counter, and When to Use Them Yourself, Effectiviology.
- TV Tropes article about The War On Straw, a fairly in-depth list of the many, many forms the strawman fallacy can show up in fiction.
- Straw Man Kit instructions
- You Can Lead an Atheist to Evidence, But You Can't Make Him Think: Answers to Questions from Angry Skeptics, p. 30.
- The Ideological Turing Test, Bryan Caplan, JUNE 20, 2011
- Ideological Turing Test - Christian vs Atheist, 2011 and 2012
- Quora: What is an ideological Turing test?
- Iron-Manning, Again, The Non Sequitur