Fallacies: Types, Detection & Avoidance in Reasoning

TAKEAWAYS:

• Fallacies are errors in reasoning that make arguments appear valid when they are not, often by exploiting emotional biases or logical gaps.
• They are classified into formal (violating logical structure), informal (flawed content), and rhetorical (persuasive but misleading) types.
• Common informal fallacies include ad hominem, straw man, false cause, and appeal to authority, each with distinct patterns.
• Real-world impact: Fallacies appear in ads (e.g., Daraz’s "everyone buys this"), politics (e.g., Ncell’s "our network is best" without data), and even academic writing.
• Detection tools: Question assumptions, check evidence, and ask "Does this hold under scrutiny?" to spot fallacies.
• Avoidance: Structure arguments with premises → conclusion clarity, cite credible sources, and test for logical consistency.

1. What Are Fallacies?

Fallacies are flaws in reasoning that make an argument seem convincing but are logically unsound. They can be:

• Intentional (manipulative rhetoric).
• Unintentional (genuine mistakes in thought).

Why study them? Fallacies undermine critical thinking, a core skill for programmers (e.g., debugging flawed algorithms), analysts (e.g., interpreting data), and citizens (e.g., evaluating news).

2. Types of Fallacies

Fallacies are categorized based on their source of error:

A. Formal Fallacies

Violate the structure of logical arguments (e.g., syllogisms). Example:

• Affirming the consequent:

  "If it’s a dog, it barks. It barks, so it’s a dog." Error: The converse of the premise is assumed to be true.

Visual: Formal Fallacy Structure

graph LR
    A["Premise 1: If P, then Q"] --> B["Premise 2: Q"]
    B --> C["Conclusion: Therefore, P"] --> D["❌ FLAWED"]

B. Informal Fallacies

Flaws in the content or context of arguments. These are more common in everyday reasoning.

1. Ad Hominem ("To the Man")

Attacking the person instead of the argument.

• Example:

  "You can’t trust his climate change report—he’s not even a scientist!"

• Real-world tie: Political debates often use this to discredit opponents (e.g., "Nepali Congress’s plan is useless because they’re corrupt").

2. Straw Man

Misrepresenting an opponent’s argument to make it easier to attack.

• Example:

  "You support free healthcare? So you want to bankrupt the country!" (Original argument: "Expand healthcare access gradually.")

• Real-world tie: Ads for Khalti often exaggerate competitors’ flaws (e.g., "Other apps charge fees—Khalti is free!").

3. False Cause (Post Hoc Ergo Propter Hoc)

Assuming correlation = causation.

• Example:

  "I wore my lucky socks, and my team won. The socks caused the victory."

• Real-world tie: NEPSE stock analysts often claim "The market rose because of X" without data (e.g., "Prime Minister’s speech boosted shares").

4. Appeal to Authority (Argumentum ad Verecundiam)

Using an unqualified authority to support a claim.

• Example:

  "A famous actor says this diet works, so it must be true."

• Real-world tie: Pathao ads feature celebrities, implying "If they use it, it’s reliable"—but celebrities lack expertise in ride-hailing.

5. Appeal to Emotion (Pathos Fallacy)

Manipulating emotions instead of logic.

• Example:

  "If you don’t support this law, children will suffer!" (Fear-mongering)

• Real-world tie: NTC ads use guilt ("Your family deserves fast internet!") to push subscriptions.

6. Slippery Slope

Claiming a small step will lead to an extreme outcome without evidence.

• Example:

  "If we allow same-sex marriage, next people will marry animals!"

• Real-world tie: Critics of eSewa’s digital payments argue "It will replace cash entirely, destroying small businesses."

7. Hasty Generalization

Drawing a broad conclusion from insufficient evidence.

• Example:

  "I met two rude Nepalis in Kathmandu, so all Nepalis are rude."

• Real-world tie: Daraz reviews often generalize ("This product is terrible!") based on one bad experience.

8. Circular Reasoning (Begging the Question)

The conclusion is assumed in the premise.

• Example:

  "The Bible is true because it says so."

• Real-world tie: Ncell claims "Our network is the best because we say so" (no third-party data).

3. Rhetorical Fallacies

Used in persuasion (e.g., marketing, politics) to sway audiences without logic.

4. How to Detect Fallacies

Use the "5-Question Test" for any argument:

1. Is the premise true? (Check evidence.)
2. Does the conclusion follow logically? (No gaps?)
3. Are there hidden assumptions? (Unstated premises?)
4. Is emotion being manipulated? (Fear, pity, guilt?)
5. Is the authority credible? (Expertise in the field?)

Example: Analyzing a Daraz Ad

"90% of buyers love this phone! Get it before it’s gone!"

• Fallacy: Bandwagon + Hasty Generalization.
• Why?
  • "90%" may ignore non-buyers.
  • "Before it’s gone" creates urgency (scarcity = emotion).

5. Avoiding Fallacies in Writing/Code

For Essays/Reports:

• Structure arguments clearly:

  Premise 1 → Premise 2 → Therefore, Conclusion.

• Cite sources (e.g., "NTC data shows...").
• Test for consistency: If your conclusion contradicts evidence, revisit premises.

For Programming/Algorithms:

• Formal fallacies = bugs in logic.
  • Example: A loop assuming "If X, then Y" without handling exceptions.
• Informal fallacies = flawed assumptions.
  • Example: A sorting algorithm that "works for small datasets" (hasty generalization).

Code Example: Avoiding False Cause

# ❌ Flawed: Assumes correlation = causation
if user_clicked_ad and purchase_made:
    print("Ad caused purchase!")  # False cause!

# ✅ Better: Track multiple variables
if (user_clicked_ad and
    purchase_made and
    user_was_logged_in and
    inventory_was_available):
    print("Possible conversion path.")

6. Real-World Applications

A. Business & Marketing

• eSewa: Avoids "Pay us or lose services" (emotional blackmail).
• Daraz: Uses "Limited stock!" (scarcity = slippery slope).
• Ncell: Claims "Best network" without comparative data (appeal to authority).

B. Politics & Media

• Nepali Congress vs. CPN-UML: Straw man attacks ("They want to destroy democracy!").
• YouTube Algorithms: Bandwagon effect ("Trending because you watched!").

C. Everyday Life

• Traffic in Kathmandu:
  • "If we ban private cars, everyone will use bikes!" (Slippery slope)
  • "My friend got stuck—traffic is terrible!" (Hasty generalization)
• Bank Loans:
  • "Take this loan—everyone gets approved!" (False cause)

7. Common Fallacies in Nepali Context

8. Worked Example: Analyzing a News Headline

Headline: "New NTC Policy Will Save 50% on Bills—Experts Agree!"

1. Fallacy Check:
   • Appeal to Authority: "Experts"—are they neutral?
   • Overgeneralization: "50% savings"—for whom? All users?
2. Critical Questions:
   • What’s the sample size of "experts"?
   • Are there hidden costs (e.g., slower speeds)?
3. Revised Claim:

   "Preliminary data from 100 users shows a 20% reduction in bills under specific conditions."

9. Visual Summary: Fallacy Detection Flowchart

flowchart TD
    A["Is the argument emotional?"] -->|"Yes"| B["Appeal to Emotion\nRed Herring"]
    A -->|"No"| C["Does it attack the person?"] -->|"Yes"| D["Ad Hominem"]
    C -->|"No"| E["Does it assume causation?"] -->|"Yes"| F["False Cause"]
    E -->|"No"| G["Is the sample too small?"] -->|"Yes"| H["Hasty Generalization"]
    G -->|"No"| I["Is the conclusion assumed?"] -->|"Yes"| J["Circular Reasoning"]
    I -->|"No"| K["No fallacy found"]

10. Exam Tip: How to Score Full Marks

1. Identify + Name the Fallacy:
   • "This is an ad hominem fallacy because..."
2. Explain Why It’s Flawed:
   • "It attacks the person (X) instead of the argument (Y)."
3. Provide a Counterargument:
   • "A valid argument would require evidence that..."
4. Use Real-World Examples:
   • "Like in Daraz ads where..."
5. Structure Your Answer:
   • Fallacy Name → Definition → Example → Why It’s Wrong → How to Fix.

Sample Answer Format:

Question: "Evaluate the following argument: ‘You can’t trust climate science—Al Gore isn’t a real scientist.’" Answer: This is an ad hominem fallacy because it attacks Al Gore’s credibility instead of addressing the scientific evidence for climate change. A valid argument would examine peer-reviewed studies, not personal attacks. For example, the IPCC reports (cited by over 130 countries) provide data-backed conclusions, unlike this emotional dismissal.

11. Practice Questions for TU/PU Exams

1. Identify the fallacy:

   "If we legalize marijuana, soon people will be smoking it in schools!" Answer: Slippery slope.

2. Debunk the fallacy:

   "Khalti is unsafe because a hacker stole money once." Answer:

   • Fallacy: Hasty generalization.
   • Counter: "Security breaches are rare; Khalti uses encryption like banks do."

3. Design a fallacy-free argument: Claim: "Nepal’s internet is slow." Valid Argument:

   "According to NTC’s 2023 report, Nepal’s average download speed is 12 Mbps, below the regional average of 25 Mbps (ADB, 2023). This suggests infrastructure upgrades are needed."

12. Key Takeaways for Exam Day

• Memorize 5 common fallacies: Ad hominem, straw man, false cause, bandwagon, circular reasoning.
• Use the "5-Question Test" to analyze any argument.
• Link to real-world examples (e.g., ads, politics, tech) to show understanding.
• Avoid vague language: Instead of "This is wrong," say "This is a false cause fallacy because...".