The Six Reasoning Strategies of Sherlock Holmes

(Following Part 1)

Holmes’s Reasoning Method Adopted by the FBI

Sherlock Holmes has transcended the image of a mere detective. He has become a symbol of logical and scientific thinking—and a subject of study for many. The U.S. Federal Bureau of Investigation (FBI) has drawn inspiration from Holmes’s techniques in developing profiling methods.

In business, many companies apply Holmes’s method of elimination and the hypothetico-deductive method to break down complex problems into smaller components to arrive at the most rational and data-driven decisions.

In the rapidly evolving field of generative artificial intelligence (AI), the technique of “prompt engineering” — formulating clear questions and hypotheses to draw accurate and creative results from AI — is also being developed based on Holmes’ logical reasoning approach.

Introduction to Holmes’ Logical Reasoning Techniques

Depending on the nature of the case and the type of clues available, he selects the most effective method—and often combines multiple techniques. The six primary methods Holmes employs are: deduction, induction, abduction, hypothetico-deductive method, retroduction, and method of elimination. In this section, we will examine each technique in detail — its basic concept and how it works in practice.

Deduction (연역법)

The English word “deduction” originates from the Latin deducere, where de- means “down” or “away,” and ducere means “to lead.” Thus, deduction means “to draw out” or “to lead down,” referring to the method of deriving a specific and clear conclusion from a widely known general principle or rule.

For example, from the universal truth, “All men are mortal,” and the specific fact, “Socrates is a man,” we can naturally deduce the clear conclusion: “Therefore, Socrates is mortal.”

Holmes carefully observes small clues at a crime scene (e.g., muddy footprints on the floor) and, based on already known general facts (such mud exists only in northern London), draws a logical conclusion: the person came from northern London.

Induction (귀납법)

The English word “induction” comes from the Latin inducere, with in- meaning “into” or “inward,” and ducere meaning “to lead.” So, induction means “to lead in” or “to draw together,” reflecting the operation of this technique — gathering individual facts or cases to arrive at a general principle or rule.

For instance, upon observing that “the swans in this place are white, the swans over there are also white, and the swans elsewhere are likewise white,” we repeatedly see individual facts and thus reach a general conclusion: “All swans are white.”

Holmes gathers recurring patterns or phenomena across several cases — for example, if in multiple incidents “the dog did not bark when the intruder entered” — he concludes: “The culprit is someone familiar with the victim’s household.”

While deduction moves from “rule to conclusion,” induction climbs from “facts to rule.”

Abduction (가추법)

The English word “abduction” is derived from the Latin abducere, with ab- meaning “away” or “apart,” and ducere meaning “to lead.” Thus, abduction means “to lead away” or “to draw outwards,” referring to a reasoning method that brings forth the most plausible hypothesis to explain a given phenomenon. It represents a thought process that proposes a hypothesis to explain an observed fact.

For example, upon seeing wet grass in the morning, one might hypothesize: “It rained overnight,” or “The sprinkler system activated,” or “Dew formed.” The purpose is to select the most plausible explanation. This approach does not confirm absolute truth but assumes the most persuasive explanation.

Holmes might encounter a scene where a victim has collapsed in front of a desk, with an unfinished letter and an extinguished lamp on top. From these clues, he formulates the most probable hypothesis: “The victim was attacked while writing a letter to someone.” This explanation most naturally accounts for the observed facts.

Although abduction does not guarantee the truth of a conclusion like deduction or induction, it is extremely useful for uncovering hidden causes behind phenomena and exploring their possibilities.

Hypothetico-Deductive Method (가설-연역법)

The term “hypothetico-deductive method” consists of two parts. “Hypothetico” comes from the Greek hypotithenai, meaning “to place under” — essentially, to assume or propose an idea. “Deductive” is the same as previously discussed. Thus, the hypothetico-deductive method involves proposing a hypothesis, predicting what should occur if the hypothesis is true, and then verifying that prediction through actual observation or experiment.

For instance, given the hypothesis “Plants cannot photosynthesize without light,” one can predict that in the absence of light, plants will not release oxygen. Conducting an experiment to confirm that plants indeed do not emit oxygen under such conditions allows us to evaluate the hypothesis’s validity.

Holmes sets up a hypothesis (e.g., “The culprit is an acquaintance of the victim”) and predicts a circumstance that should be observed if the hypothesis is correct (e.g., “The victim’s dog didn’t bark at the intruder”). He then confirms this through observation — the dog did not bark — thereby validating the hypothesis.

The hypothetico-deductive method uses deductive logic to draw specific conclusions, but begins with a hypothesis, whose validity is confirmed through empirical means.

Retroduction (귀추법)

The English word retroduction comes from the Latin retroducere, where retro- means “backward” or “in reverse,” and ducere means “to lead.” In other words, retroduction literally means “to lead backward” or “to bring something back into the past.” It refers to a method of reasoning in which one starts with an observed result or phenomenon and works backward to infer the possible causes or circumstances that led to it. This form of reasoning is often described as a kind of “logical time travel.”

For example, imagine finding a broken cup on the table in the morning while knowing that no one was home. Based on this result, one might propose various possible causes: “The cat knocked it over,” or “Someone must have dropped it.” From there, you work backward through possible explanations—such as, “There was no one home, but the cat could have jumped on the table,” and, “The cat’s food bowl was empty last night, which could have prompted it to behave unusually.” Ultimately, you reconstruct the situation by narrowing down the most plausible explanation.

Sherlock Holmes uses this method frequently when analyzing crime scenes that present closed, seemingly inexplicable circumstances. For example, upon discovering a body in a locked room—with windows unbroken and doors locked from the inside—Holmes looks at these final conditions and reasons backward to infer their causes. He may conclude, “This incident was not the work of an intruder, but rather someone already inside the room.” He then builds his investigation around that retroductive hypothesis.

Though similar to abduction (as previously described), retroduction differs in direction and emphasis. While abduction focuses on selecting the most plausible explanation from among several hypotheses to explain an observed phenomenon in the present, retroduction emphasizes tracing backward from a known outcome to infer what might have led to it.

Retroduction is particularly powerful for analyzing past events or deducing motives and methods of a crime—factors that are often unobservable and must be logically reconstructed after the fact.

Method of Elimination (제거법)

The method of elimination is a logical reasoning process where one systematically eliminates all options that are inconsistent with the facts or impossible, leaving only the most plausible answer—regardless of how improbable it may initially seem.

For instance, when investigating a case with multiple suspects, Holmes might eliminate those with clear alibis one by one. In the end, if only one person remains without a verified alibi, that individual becomes the prime suspect. Holmes’s well-known quote, “When you have eliminated the impossible, whatever remains, however improbable, must be the truth,” captures the essence of this method perfectly.

The strength of this method lies in the fact that you do not need to know the correct answer from the beginning. As long as you’re able to logically eliminate incorrect possibilities, you can narrow down to the truth.

What sets Holmes apart is not just his mastery of these methods individually, but his genius in combining them seamlessly depending on the situation. Each method is powerful on its own, but when used in combination, they become even more formidable tools of logic and deduction. Holmes’s unique approach comes from his fluency in switching between these methods and tailoring his reasoning to the situation at hand.

In the next article, we’ll explore how Holmes actually applies these reasoning methods in combination throughout his most iconic cases from the original stories.

About the Author:

Park Seung-ryong has worked in journalism, PR agencies, the Presidential Office, and cultural content institutions, consistently nurturing a deep interest in logical and insightful thinking. He firmly believes that Sherlock Holmes’s reasoning methods are invaluable for cultivating “thinking power” in the era of generative AI. After retirement, he has been actively promoting the “Sherlockian Way of Thinking.”