Site Homepage » Logic » Informal Fallacies » Fallacy of Accident

The Fallacy of Accident: Fallacia Accidentis

1. The Fallacy of Accident Defined

   
   
   1. Fallacy of Accident: the logical mistake of inferring an atypical specific example or examples from a general premise which does not apply to that instance or those instances.
      
      This fallacy occurs whenever anyone concludes that the scope of a generalization refers to uncharacteristic or incongruous examples of that generalization.
      
      

      ---

      Caveat for the Fallacy of Accident: Historically, the fallacy of accident has been defined in several discrepant ways under different designations; consequently, some recent informal logicians conclude the fallacy should not be covered in the standard treatment of informal fallacies. Q.v., Postscript below.

      Since the time of Aristotle, the fallacy of accident has been occasionally identified with the fallacies of fallacia accidentis, a dicto simpliciter ad dictum secundum quid (arguing from a statement made generally to a qualified statement), converse accident, and a dicto secundum quid ad dictum simpliciter, (arguing from a qualified statement to a statement made generally), among others. The noted informal logician Douglas Walton recommends combining all the fallacies in this class under any one of the following names: “the neglect of qualifications,” secundum quid (i.e., somehow qualified), or “overlooking special circumstances.”[2]

      ---

      
      
      
   2. All things are generally classified in accordance with the general features of resemblance among their instances. Individual things of this kind also, of course, can differ in other non-characteristic respects as well. The fallacy of accident occurs when a dissimilar respect (or accidental feature) of some kind is mistakenly claimed to be a consequence of being of that kind.

      E.g., This type of mistake occurs if someone were to argue that since everyone in the U.S. has the right to vote, children, criminals, and noncitizens have the right to vote.

      Thus, the claim that accidental properties generally true of something are always true of each and every individual of that type is a fallacious claim. This error is sometimes explicitly prevented by adding a qualifying phrase to these kinds of non-universal generalizations, such as in statements like these:

      “As a rule, such and such is true.”
      “Normally, such and such is true.”
      “Typically, such and such is true.”
      “On the whole, such and such is true.”
      “In general, such and such is true.”
      

      
      Abstract principles and rules of thumb have to allow for qualifying conditions. Sometimes the accidental feature of a kind or type of thing is said to be “the exception that proves the rule.”
      
      
      
      1. Hermann Lotze cites this fallacy as characteristic of a demagogue:

         “The doctrinaire is an idealist, who refuses to see that though ideas may be right in the abstract, yet the nature of the circumstances under which and of the objects to which they are to be applied must limit not only their practicability but even their binding force.”[3]

         For example, the fallacy would occur if a disciplinarian were to claim that there are no exceptions to the general ethical rule that one ought never lie; therefore it is wrong to write poetry or fiction, tell a joke, or respond agreeably to awkward questions.
         
         
         Historically, the fallacy of accident was first described in English in 1599 by the Renaissance polymath Thomas Blundeville in his Arte of Logike:

         ”Fallacia Accidentis … whch may be Englished thus: The Fallax of the Accident …”[4]

         Blundeville describes the fallacy, heretofore termed fallacia accidentis, as comprising of both accident and converse accident.
         
         
         
      2. A century ago, logician H.W.B. Joseph described the type of non-universal generalization used in many examples of the fallacy of accident in this manner:

         “[T]here is no fallacy more insidious than that of treating a statement which in many connexions is not misleading as if it were true always and without qualification.”[5]

         More recently, the pragmatic informal logic theorist Douglas Walton writes:

         “The default type of generalization can also be called the defeasible

         close ×

         Defeasible:

         Capable of being overturned … A proposition is defeasible if further evidence may render it doubtful. Simon Blackburn, The Oxford Dictionary of Philosophy (Oxford: Oxford University Press, 1994), 96.

         (presumptive) generalization, a type of generalization that admits of exceptions, and is compatible with some new arguments that turn up counter-instances. This type of generalization is not strict, but is open ended and tentative in nature.”[6]

         Examples of these types of generalizations appearing in the fallacy of accident include include abstract principles[7], rules of thumb, legal and moral laws, maxims, adages, clichés, aphorisms, proverbs, and limited generalizations.
         
         
         
      3. These types of generalizations (i.e., those which are true under certain circumstances) are sometimes described in philosophy and logic as “soft generalizations.” Examples are provided in Section III Explained Examples below.
         
         In sum, soft generalizations (defeasible generalizations) need to allow for qualifying conditions.
         
         
         
      4. S.H. Emmons, among many other 19th century logicians, recognizes the subtleties of the fallacy of accident:

         “Fallacies of this nature are very difficult to detect, in consequence of the readiness which exists to lose sight of the limitations which alone render a general law capable of being applied to some particular case.… [T]here are many various ways of sophistication by means of ambiguity in expression which … are often employed.[8]

         Thus, “accidental” or non-essential instances of a generalization can differ in virtually limitless ways. The noted argumentation theorist Douglas Walton goes so far as to conclude:

         “Whatever accident is, it is far from clear that it is the kind of common error of reasoning or fallacy that belongs in the standard treatment of fallacies in logic textbooks …”[9]

         Indeed, many current logic textbooks omit the treatment of the fallacy of accident altogether.
         
         
         
   3. Generally speaking, historically important definitions of the fallacy of accident have shifted from the fallacy defined in the use of terms, for example …

      “Socrates is white.
      White is a color.
      Therefore, Socrates is a color.”[10]

      … to the fallacy of accident defined in terms of the use of statements, for example:

      By law, all children must attend school.
      ∴ Your child, although ill, must attend school.

      The definition in terms of statements is often equated with the fallacy of a dicto simpliciter ad dictum secundum quid (i.e.. the fallacy of arguing from what is true absolutely to what is true in a certain [qualified or accidental] respect).
      
      From the point of view of definition, the fallacy of accident is often said to have the same extension (i.e. refers to the same things) as this kind of secundum quid but have a different intension (i.e. different described properties). Accident is defined by its use of terms, whereas usually secundum quid is defined by its relation of statements.
      
      
      
      1. The initial phrasings of the fallacy of accident from which the contemporary versions are derived were due to Aristotle:

         “Fallacies connected with Accident occur when it is claimed that some attribute belongs similarly to the thing and to its accident; for since the same thing has many accidents, it does not necessarily follow that all the same attributes belong to all the predicates of a thing and to that of which they are predicated.” [Aristotle, Soph. El. 166b 28-33.][11]

         Even though Aristotle considered accident to be a material (i.e. non-verbal) fallacy, the fallacy of accident during the medieval period was regarded, for the most part, as a verbal fallacy. E.g. John Wallis summarized:

         “The fallacy of accident is when something accidental is confused with what is primarily intended to be essential or principally intended.”[12]

         One way Wallis's definition was interpreted is that a term used or supposed materially (or referentially) should not be confused with a term used formally (non-materially or abstractly).
         
         For example in the sentence “The cat is a species” the subject is used formally, whereas in the sentence “The cat is a calico” the subject is supposed in a material sense.
         
         When the supposition of a term is changed in the midst of an argument, the fallacy of accident is said to occur because the change is in effect a change in the designation of the term.[13]
         
         
         
      2. The following definition of the fallacy is a restatement by Henry Aldrich in his Compendium of Logic (1756) — a definition repeated in some logic texts until the end of the 18th century:

         “The Fallacy of the Accident; when some accidental Circumstance is confounded with what is essential.”[14]

         Aldrich provides the following example — an example commonly used in logic textbooks until the early 20th century:

         What destroys Men ought to be prohibited.
         Wine destroys Men.
         Therefore Wine ought to be prohibited.[15]

         Essential characteristics of objects or events are said to be always present in objects or events of that kind; accidental characteristics are said to be not always present. Aldrich explains his example:

         “The major Proposition must mean, What necessarily destroys Men: Otherwise it is not true: The minor, Wine accidentally destroys Men. Therefore here also there are four Terms.“[16]

         Note that in this example of the fallacy, all statements are general statements, but the major premise is more general than the general instances named in the conclusion. Note also that Aldrich points out that the [informal] fallacy of accident in this example also commits the syllogistic formal fallacy of the four term fallacy.
         
         
         
      3. In his 19th century celebrated textbook, W. Stanley Jevons defines the fallacy in modern terms:

         “The fallacy of accident consists in arguing erroneously from a general rule to a special case, where a certain accidental circumstance renders the rule inapplicable.”[17]

         Jevons’s formulation of the fallacy was adopted by a great many textbooks of the period, and his definition appears frequently in logic textbooks and on webpages today.
         
         
         
      4. In the latter 20th century, many informal logicians concurred with Douglas Walton's assessment that the fallacy of the accident has had such a checkered history that if it is to be taught at all in logic classes, it's best to be thought of as a ”fallacy of ignoring exceptions.“

         ”Trying to explain the historical origins of unfamiliar and puzzling terms like ‘accident’ … seems a hopeless and unrewarding task, not to mention the variety of Latin terms and phrases peppered through the traditional textbook treatments. While ignoring exceptions to a rule or qualifications to a generalization did seem important failures to warn students about, nevertheless the complications and puzzles inherent in the standard treatment suggested that it was prudent to bypass this area, restricting coverage to a brief mention of the basic fault, in simple terms.”[18]

         In general, the fallacy of accident described in textbooks today is illustrated by examples arguing from general to specific, genus to species, essence to accident or abstract to concrete terms — all cases where the existence of qualifying circumstances are neglected.
         
         The examples of the fallacy of accident on this website in notes, quizzes, and tests are restricted to the fallacy of ignoring exceptions to a rule or ignoring qualifications to non-universal generalizations.
         
         
      5. A historically exemplary instance of ignoring qualifications or secundum quid which also exhibits the characteristics of arguing erroneously from a general rule to a special case where a certain uniquie (accidental) circumstance renders the rule inapplicable is as follows:

         Water is liquid.
         Ice is water.
         ∴ Ice is liquid.

         In the first (major) premise, the term “water” is used in accordance with its essential definition in a general sense without condition; however, in the second (minor) premise, the term is used to refer to water in the conditional sense of being frozen. So the premises shift from a basic, unconditional statement to a statement incorporating a specific condition.

         From another point of view, the argument proceeds from a general rule (“Water is liquid) to conclude with a special case where the subject is restricted to the unrepresentative case of water in a certain phase, namely that of ice.
         
         
         

2. The Fallacy of Accident Argumentation Scheme

   Informal Guide to the Fallacy of Accident

   A generalization or rule is normally true.

   ∴ An exceptional instance of that generalization is claimed true.

   
   The “accidential” features of the specific case make it an exception to the underlying conditions of the generalization or rule. The conclusion of the argument is not an instance commensurate with all typical examples included in the meaning of the major premise.
   
   
   
   1. E.g., the following argument is often used as an example of the accident fallacy:

      Water is liquid.
      Ice is water.
      Ice is liquid.

      “Water” is used in its unexceptional, unconditioned meaning in the premises; however, in the conclusion “water“ is used in an atypical sense. The condition of being ice is not essential to being water.
      
      
      
   2. Another way this scheme is ordinarily understood is that the fallacy of accident occurs in arguments where a qualified or atypical statement is concluded from an ostensibly absolute statement.
      
      Most generalizations and rules have exceptions, and to treat these rules as universally true absolute generalizations is faulty reasoning.
      
      
      

3. Some Common Varieties of the Fallacy of Accident

   Some of the common types of the fallacy of accident are illustrated in this section.
   
   
   
   1. From a Soft Generalization: Soft generalizations are statements claiming something is usually, mostly, typically, or normally the case. So the general statement is ordinarily true but can have exceptions for any number of different reasons. Types of soft generalizations include abstract principles, rules of thumb, legal and moral laws, and aphorisms.
      
      
      
      1. Soft generalizations are the basis of most examples of the accident fallacy provided in current logic textbooks. Here are two examples of a soft generalization from traditional textbooks:

         “All persons who do not pay import duties are smugglers. (Envoys have immunity.)
         All students off grounds are truant. (Some have permits.)”[19]

         In the statement of the generalization, exceptions are normally understood.
         
         
         
      2. Another illustration of this kind of fallacy is the following tacit inference from the soft generalization “On country walks, there is much to be seen.”

         “During a country walk there is much to be seen: but a country walk may be taken on a pitch-dark night: therefore on a pitch-dark night there may be much to be seen.”[20]

         In this example, the presumed time condition of country walks during the daytime is ignored in the conclusion to the argument.
         
         
         
      3. H.H. Munro offers this example:

         “Fidelity is a virtue.
         Allegiance to an ignoble cause is fidelity.
         ∴ Allegiance to an ignoble cause if a virtue.”[21]

         “Allegiance to an ignoble cause” is not, as such, representative of “fidelity” in general. The more specific use of the concrete “accidental” term “fidelity” in the minor premise is used in contrast to a more general use of the term in the major premise — the contrasting usages of the term give rise to the fallacy of accident.
         
         
         
      4. Another fallacy which uses a general statement having an implicit qualified meaning as if it were not so qualified is illustrated in the following argument:

         E.g., “[A]s if a traitor should argue from the sixth commandment, Thou shalt not kill, to prove that he himself ought not be hanged.”[22]

         The premise “Thou shalt not kill” treated as a generalization which refers to the unlawful or unjustified killing of human beings, and its meaning is misconstrued in the conclusion to refer to a different specific example (presumably a locally lawful or military judicially justified instance) from those denoted in the premise.
         
         This argument can be interpreted as follows:

         No persons should kill other people.

         ∴ No persons should kill me, a traitor.

         Note two important things:
         
         
         
         1. The grammatical form of the premise “Thou shalt not kill” is stated as a particular premise, but in the context of being a religious moral law, the premise is intended to be a statement applying to everyone:

            “No persons should kill [in the sense of “unlawful killing] other persons [all other things being equal].”

            So, sometimes, statements which appear to be particular statements are intended to be understood as general statements.
            
            
            
         2. Even if the conclusion were to be stated as a general statement, e.g. …

            “No persons should kill traitors.”

            … the argument would be the fallacy of accident. Note this general conclusion is narrower in scope that its premise:

            “ No persons should kill other persons.”

            The fallacy of accident concludes with an atypical instance said to have been implied by a generalization which is not meant to include that instance. In this argument, the conclusion, itself, is a generalization, but the generalization is about individuals of a different sort that the generalized premise.
            
            So, it's important to notice that the atypical instance mentioned in the conclusion can be, and often is, a generalization about circumstances of less scope than the premise.
            
            
            
         3. Finally, it's important to notice that even if both premise and conclusion were to be expressed as particulars, the argument would be an accident fallacy:

            One should never kill another person.
            An executioner should never kill a legally condemned person.

            This example illustrates why the accident fallacy is best identified as a fallacy of neglecting qualifications rather than the inference of an exceptional case from a general rule.
            
            
            
         4. Consequently, many cases of the fallacy of accident cannot be analyzed in terms of generality or specificity. Indeed, traditionally, many logicians do not distinguish among the fallacies of accident, converse accident, and secundum quid.
            
            
            
   2. From Moral and Legal Principles: Moral and legal rules are to be understood in terms of “all other things being equal.” Interpreting a moral or legal law as an absolutely universal statement or hard generalization can lead to the fallacy of accident.
      
      This variety of the fallacy of accident can occur when a judge misinterprets the application of a moral or legal rule. Jevons points out the ubiquity of the fallacy of accident in legal issues:

      “Almost all the difficulties which we meet in matters of law and moral duty arise from the impossibility of always ascertaining exactly to what cases a legal or moral rule does or does not extend; hence the interminable differences of opinion, even among the judges of the land.”[23]

      Additionally, whenever the wrong legal or moral standard is followed an error of law is said to occur.
      
      
      
      1. E.g., Felicia Hemans' well-known poem Casabianca recounts an actual incident from the 1798 Battle of the Nile concerning a young boy who refused to leave a burning ship without his father's orders (interpreted as a moral rule):
         
         

         “The boy stood on the burning deck

            Whence all but he had fled;

         The flame that lit the battle's wreck,

            Shone round him o'er the dead.

         …
         
         

         The flames roll'd on—he would not go

            Without his Father's word;

         That Father, faint in death below,

            His voice no longer heard.”[24]

         Casabianca's father never intended his son to remain on deck under such dire circumstances. In some unanticipated circumstances, prescriptive rules have exceptions.
         
         
         
      2. In this next example, a contemporary editorialist asserts socialistic left-wing politicians violate a religious commandment:

         “The Ten Commandments prohibit theft and envy, the pillars of socialism — using political power to seize and redistribute private property — which the highly politicized left-wing crowd so loves.”[25]

         Since the proper application of the Ten Commandments as religious law is being applied outside its sphere of application to democratically derived consensual governmental law, the fallacy of accident occurs.
         
         
         
      3. Whenever a prescription is invoked away from its frame of reference and is pressed beyond its sphere of application, the fallacy of accident will occur.
         
         
         
   3. From the First Use of a Term Rendered Abstract and the Second Use Concrete: Abstract terms refer to immaterial things (usually general and qualitative) such as qualities, concepts, states, events, and ideas which as a rule cannot be sensed, whereas concrete terms usually refer to material (usually specific and quantitative) or tangible objects which can be sensed.
      
      
      
      1. The difference between an abstract and a concrete term is not polar relationship but is a matter of degree: many abstract and concrete terms exist on a continuum and can be abstract or concrete depending on their context in a passage.
         
         
         
      2. E.g. the abstract term “virtue” can, in context, be used concretely if it is used as separable into different kinds: the virtues of courage, justice, prudence, and so forth. So in a sentence such as …:

         “Courage is the price that life exacts for granting peace,”[26]

         … “courage” is used more concretely than in the sentence:

         “Courage doesn't always roar.”[27]

         However, identifying the accident fallacy solely in terms of the abstract-concrete distinction is only one variety of the fallacy.
         
         
         
      3. Consider the following syllogism devised by Thomas Crumley:

         “Fidelity is a virtue.”
         Allegiance to an ignoble cause is fidelity.
         ∴ Allegiance to an ignoble cause is a virtue.[28]

         The more specific “accidental” concrete term “fidelity” in the minor premise is used in contrast to a more general use of the term in the major premise — the contrasting usages of the term gives rise to the fallacy of accident.
         
         
         
   4. In Syllogisms as Equivocation or the Four Term Fallacy: To reason from a hard generalization to a particular instance covered by that generalization is one kind of deduction and is ordinarily necessarily valid unless ambiguity of a term is present in the argument.
      
      Historically, the fallacy of accident has been described as an ambiguous middle term in a syllogism.[29] Nevertheless, any of the three terms in a syllogism might have this ambiguity, although it is often the middle term (the term in both premises) when an ambiguity is present. The fallacy of accident often occurs when one of the terms is used in a general or definitional sense (i.e., its essential sense or essence) in one premise and a difference sense of the term is used in a special, limited, or restricted sense (i.e., an accidental sense).
      
      Thus, the fallacy of accident as equivocation in a syllogism can occur whenever a term is in a general sense in one premise and in an atypical particular or less general sense in the other premise. When considered as a syllogistic argument, such fallacy can be regarded alternately as an informal fallacy or a formal one since, in a sense, the proper form of the syllogism is broken because essentially more than three terms are present.
      
      
      
      1. E.g., the following argument appears at first glance to be a valid-formed argument but is not valid:

         Education develops the subject's intellect.
         The training of animals is education.
         ∴ The training of animals develops their intellect.[30]

         The common term “education” in each premise is used in difference senses. Some types of education do not develop the intellect; the training of animals is consider education only in a Pickwickian sense.
         
         Some informal logicians who see the major premise as a soft generalization and the minor premise as an exception to that generalization, label this argument the fallacy of accident (or one form of secundum quid). However, the mistake in the argument is best seen as the fallacy of equivocation (or, formally speaking, four term fallacy).
         
         
         
      2. In 1686, Henry Aldrich repeats the oft-used straightforward example (used above) of a syllogistic four-term fallacy committing fallacia accidentis:

         “What destroys Men ought to be prohibited.
         Wine destroys Men.
         Therefore Wine ought to be prohibited.”[31]

         Aldrich like John Wallis[32] before him described this fallacy as “some accidental circumstance is confounded with what is essential.”
         
         Most historical references (1600-1850) used this same definitional description for the more general fallacy of fallacia accidentis and often did not distinguish the fallacy of converse accident from that of accident. However, in this example when the major premise is seen a soft generalization and the conclusion is viewed as an “accidental” instance of that generalization, the fallacy was considered to be that of accident.
         
         Aldrich argues that since the first premise denotes what “necessarily destroys men” and the second premise denotes what “accidentally destroy men,” the fallacy of accident occurs.
         
         
         
      3. F.C.S. Schiller provides a subtler example of the fallacy of accident occurring in a syllogistic four-term fallacy:

         “[I]t may be generally true that ‘all men love good stories,” and undeniable that ‘Smith is a man’; yet the inference that ‘therefore Smith loves this good story,’ may be falsified in this particular case by the fact that the story is told about him, and that, therefore, he hates it.”[33]

         The ambiguity of the middle term is claimed to be based on the fallacy of accident in that Smith is a man who loves ‘good stories,’ but not a man who loves ‘good stories about him.’
         
         Other logicians have noted the occurrence of the fallacy in a syllogism when the middle term is used in one premise as applying to a general term as to its essence or general meaning (i.e., simply) and in the other premise applying to a particular, qualified, or special case of the term as accidental.
         
         Schiller goes on to say “[i]t is arbitrary whether the ambiguity of the middle [term] is described as a ‘fallacy of accident,’ or the form of the syllogism is itself regarded as involving this fallacy …”[34]
         
         Interestingly, elsewhere, Schiller suggests that no syllogism can produce certainty, since all syllogisms might commit the fallacy of accident as the fallacy of the ambiguous middle term. In the context of the statements in which it occurs, the middle term might always imperceptibly vary to some degree in its signification.[35]
         
         
         
   5. In the Immediate Inference, Conversion Per Accidens: Historically, the fallacy of accident has, at times, been based on the assumption that a term and its necessary or essential predicates are exchangeable, whereas a term and its accidental predicates are not.[36]
      
      Thomas Crumley writes:

      “[I]n effect the fallacy of Accident is based on the assumption that an accident is commensurate with its subject, and that consequently an ordinary A proposition is simply convertable. As a false principle of reasoning the assumption may be given the following expression: What is predicable of a subject is predicable of its accidents.[37]

      Crumley's use of the word “accident” indicates, for example, that whenever the subject of a statement is distributed (i.e., refers to each and every member of the subject class) and the predicate is undistributed (i.e., refers to only some of the members of the predicate class), the predicate is an accident.
      
      Nevertheless, the subject and predicate terms must be more than essential in order for valid exchange of subject and predicate of a statement to be valid. That is, the subject and predicate must be commensurate: the same intension if both subject and predicate: i.e. terms in identical statements or definitions.
      
      
      
      1. On the one hand, if a man with a brown hat steals a purse and a suspect is found who is wearing a brown hat, then the suspect with the brown hat is not thereby guilty. The thief with the brown hat is not necessarily the suspect with the brown hat.[38]
         
         
         
      2. On the other hand, a statement such as …

         All lodestones are magnetic iron ores.

         can be validly converted to …

         All magnetic iron ores are lodestones.

         because a lodestone is defined as magnetic iron ore. A lodestone is a chunk of magnetite which was magnetized, perhaps from a lightning strike. Nevertheless, in spite of its name, not all pieces of magnetite are magnetic. So, the statement …

         All magnetic iron ores are magnetites.

         is true but its converse is not true:

         All magnetites are magnetic iron ores.

         Consequently, an invalid conversion of an A type of categorical proposition (“All S is P”) is sometimes said to commit the fallacy of accident when the S and P terms are not necessarily commensurate or essential to each other, as they are not in this case of relating magnetite to magnetic iron ore.
         
         However, a valid conversion in a case like this can be made if we reduce the quantity of the statement:

         Some magnetites are magnetic iron ores.

         This inference is usually termed “conversion by limitation” or “conversion per accidens.”
         
         Nevertheless, to conclude that the universal conversion of a universal non-necessary statement (e.g. not a definition, a tautology) is a fallacy of accident extends the scope of that fallacy beyond most traditional definitions of trying to prove an special case under exceptional circumstances from a generalization.
         
         Although all planets are celestial bodies, we cannot infer all celestial bodies are planets because being a celestial body is (although necessary) is not sufficient for being a planet.
         
         
         
      3. The following example is from a philosophical dispute: F.C.S. Schiller accused formal logicians of this kind of invalid conversion when they claimed that pragmatists maintain:

         “Whatever works is true, because … whatever is true must work.”[39]

         If the class of “whatever is true” is larger than the class of “whatever works,” then the fallacy of accident might be said to occur because from the non-universal generalization “Whatever works is true,” it's also usually understood that some things that work might not be true.
         
         Nevertheless, if “whatever works” is understood in the context of being defined as “whatever is true,” then no fallacy would occur.
         
         
         
   6. Aphorisms and proverbs sometimes fashion the basis of a fallacy of accident. An aphorism is a concise statement of a general truth of good advice which usually originates from a classical or traditional source, and a proverb is a succinct saying, often metaphorical, which expresses a general truth or word of advice.
      
      
      
      1. Voltaire's aphorism …

         “The best is the enemy of the good”[40]

         assures that over-striving for achievement can block the possibility of obtaining something good enough. The fallacy of accident occurs whenever someone argues, for example, from this aphorism to the conclusion that a surgeon need not try try to do her best because she might botch her surgical procedure.
         
         It's worth noticing almost any aphorism is a non-universal statement, and usually can be opposed by some other aphorism with which it conflicts. In the case of Voltaire's aphorism, for example, the novelist W. Somerset Maugham writes, “[I]f you refuse to accept anything but the best, you very often get it.”[41]
         
         

         Some additional examples of conflicting aphorisms include:

         Absence makes the heart grow fonder.	Out of sight, out of mind.
         Look before you leap.	He who hesitates is lost.
         Do unto others as you would have others do unto you.	Nice guys finish last.

         
         
         
      2. An obvious and rather silly example of this kind of fallacy can be fashioned from a well-known Scottish proverb:[42]:

         The proverb “Kindness begets kindness” implies termites are allowable in residences.”

         The fallacy of accident occurs here because termites in houses are not the kinds of circumstances implied by the proverb; termites in houses are irrelevant to the kinds of things kindness is claimed to bring forth.
         
         

4. Arguments Sometimes Confused with the Fallacy of Accident

   1. Arguing From a Hard Generalization To a Relevant Specific Instance: If a particular conclusion is a proper instance or example of the generalized premise, no fallacy is committed.
      
      In the fallacy of accident, a general premise proceeds to a more particular conclusion where the particular conclusion is an abnormal instance, example, or case in point of the premise.
      
      
      
      1. The following example illustrates the mistake of identifying the fallacy of accident as arguing from a general rule to a special case.
         
         The late 19th century Oxford Handbook of Logic defines the fallacy of accident in this manner and provides the following example:

         “The Divine law orders that kings should be honoured: Louis Quatorze is a king: therefore the Divine law orders that Louis Quatorze should be honoured. Fallacy of Accident — arguing from general rule to special case.”[43]

         Louis Quatorze refers to the Sun King, Louis XIV of France.
         
         
         
      2. The following argument from the Roman Stoic philosopher Seneca appears to proceed logically from a general premise to a specific instance, but the argument is not a fallacy of accident:

         “[The] fact that all men agree upon something is a proof of its truth. For instance, we infer that the gods exist, for this reason, … that there is implanted in everyone an idea concerning deity …”[44]

         A summary argument of this passage can be translated as follows:

         Whatever all people agree on is true.
         All people agree on the existence of gods.
         ∴ People's belief on the existence of gods is true.

         The conclusion of this argument is not an exceptional instance of the generalization as in the accident fallacy; instead, the informal fallacy is the ad populum fallacy: i.e., simply because most people believe something is true, that truth does not logically prove the belief is thereby correct.
         
         (Note, however, Seneca's argument is unsound since the major premise “Whatever all people agree on is true” is false.)
         
         
         
   2. The informal fallacy of division is sometimes mistaken for the fallacy of accident: other times, in particular contexts, the fallacy of division is virtually indistinguishable from the fallacy of division. In the fallacy of division, a premise describing the characteristics of something considered as a whole is mistakenly claimed to imply the characteristics of its part or parts are the same.
      
      
      
      1. The well-known 19th century logician and mathematician Augustus DeMorgan recounts a law case which he claims illustrates the fallacy of accident:

         “Some years ago, a man was tried for stealing a ham, and was acquitted upon the ground that what was proved against him was that he had stolen [only] a portion of ham.”[45]

         In this case, the ambiguity of the word “ham” is based on the reference first to the whole (considered as “essential”) and next to the part of a ham (considered as “accidental.”).
         
         
         
      2. Sam Blows's 1899 exercises on the fallacies of presumption provide the following argument which is intended to be an example of the fallacy of accident:

         The verdict of the jury is seldom wrong.
         A is a member of the jury.
         ∴ His verdict is seldom wrong.”[46]

         However, in this case, the fallacy does not turn on the accidental use of the notion of a verdict, but, instead, results from the presumption that whatever is true of the jury as a whole is also necessarily necessarily true of each member of the jury as a part. The fallacy of division occurs (whatever is true for a whole of something is also true for its parts). Here the collective use of the term “verdict” in this argument is conflated with its distributed use.
         
         
         
      3. N.K. Davis gives this example of accident:

         “The Greeks produced masterpieces of art;
         The Spartans were Greeks;
         ∴ The Spartans produced masterpieces of art.”[47]

         “Greeks” is used in a generic sense in major premise and in a more particular or more specific sense in the conclusion. Davis states:

         It is fallaciously assumed that whatever is attributable to the genus as such, may be attributed to an accidental member.[48]

         Note how this example closely resembles the fallacy of division. But the class “Greeks” is not taken to mean “all Greeks without exception in the first premise.” That premise is a soft generalization. So the claim in this argument is not an example of the kind of fallacious reasoning where a class having a certain characteristic implies its members must also have that characteristic.
         
         
         
      4. The distinction between “accidental” and “essential” or “essence” has many difficulties. Current treatments of the fallacy of accident are in accord with Hamblin's assessment:

         “ Unfortunately it is often difficult to say of a property whether it is ‘essential’ or not, and few people these days would be prepared to go so far as to maintain generally that that the essential properties of every kind of thing can be uniquely specified. Yet all the most common examples depend on this assumption.”[49]

         Usually current textbooks of the fallacy of accident eschew the doctrinal interpretation of “accident” and express the fallacy in terms of generality and specificity (I.e. taking the secundum quid route). Consider the following example:

         Governments [in general] provide public goods for its citizens.
         A brutal despotism is a government [in particular].
         A brutal despotism provides public goods for its citizens.[50]

         The conclusion that a brutal despotism provides public goods for its citizens does not logically follow from the premise that governments provide public goods since the premise is a a soft generalization, i.e., one which is true for the most part.
         
         
         
   3. Converse Accident is sometimes confused with the Fallacy of Accident.
      
      Converse accident is often characterized as a fallacy that is the opposite of the fallacy of accident, and is defined as the fallacy of considering too few cases or of considering certain exceptional atypical cases and inductively generalizing to a rule that fits them alone, whereas in accident the argument proceeds from a generalization to a nontypical particular instance.
      
      
      
      1. E.g., Cusack's Principles of Logic inexplicably provides the following mistaken example of converse accident:

         “ [An] example of the converse fallacy of accident is —
         
         He who hurts another deserves punishment.
         The teacher who punishes a refractory pupil hurts another.
         ∴ Such teacher should be punished.”[51]

         The general principle in the first premise refers to instances of anyone who hurts another maliciously, whereas, in second (minor) premise refers to an atypical instance: a teacher who hurts a pupil without malice. This example clearly fits the schema for the fallacy of accident, not converse accident.
         
         
         
      2. Nevertheless, on occasion, the fallacies of accident and converse accident can be virtually indistinguishable. C.L. Hamlin, for example, notes the fallacy of accident and its usual converse are not distinguisable in the absence of argument indicators:

         “[I]n view of the ease with which an argument may be cast into different forms, the distinction is a little artificial.”[52]

         For example, consider the following two statements as premise and conclusion in an argument: (1) Drugs can lead to harmful health issues and (2) Drug use is risky. Without premise or conclusion indicators we cannot guess which is the qualified one or which statement is the conclusion:

         Drugs can lead to harmful health issues.
         Drug use is risky.
         

         OR

         Drug use is risky.
         Drugs can lead to harmful health issues.

         Indeed, both statements can be viewed as qualified or atypical.
         
         
         
      3. A more recent legal example of the confusion of accident and converse accident cropped up in a class action suit based on conditions in Albuquerque, N.M. jails where the Court of Appeals Judge Gorsuch rejected the defense's attempted application of a precedent case:

         “[I]t doesn't follow from this fact [from Midland Asphalt Corp. v. United States] that every case with a final judgment in it is appealable. The defendants are too quick to generalize and in doing so run afoul of the logical fallacy of accident. … [J]ust because final judgments can be appealed doesn't mean everything in every case that has a final judgment may be appealed.”[53]

         And by implication the fact that final judgments can be appealed doesn't mean the case at hand (i.e., the accidental instance) may be appealed.
         
         
         
      4. As noted earlier, unfortunately, many times it is difficult to determine the order of generality and specificity of premises. Some fallacies of direct (or sometimes dubbed specific) accident are non-general in both premises.
         
         Other times it is difficult to distinguish generality and specificity of premises because a term is being used in two different accidental senses. William Stanley Jevons's influential textbook (widely used from 1870-1920) provides the following example of arguing from one special case to another special case:

         “He who thrusts a knife into another person should be punished; a surgeon in operating does so; therefore he should be punished.”[54]

         Crumley also provides an example of a fallacy of accident when arguing from special case to special case.

         “Giving money to the poor encourages idleness.
         Giving money to the poor is charity.
         ∴ Charity encourages idleness”[55]

         Crumley points out the middle term is used in an “accidental sense” in both premises. In one sense, money is being given to the shiftless poor; in the other, money is being given to the deserving poor.
         
         

         Most logicians, following Jevons and, later, Crumley, separate the following fallacies:

         (1) simple accident (reasoning from generalization to exception)

         (2) specific accident (reasoning from unique case to a different unique case), and

         (3) converse accident (reasoning from exception to generalization)[56]

         
         Even so, Jevons admits:

         “[M]uch difficulty is often encountered in saying to which of the three [types of the accident fallacy] any particular example is best referred.”[57]

         These three types of accident were still listed in logic textbooks in the middle to late twentieth century.
         
         
         
      5. More recently, Douglas Walton sums up:

         “Whatever you call the fallacy or fallacies — secundum quid, accident, … and so forth — it does seem to make same sense to observe that the inference contained in it can go either way — from the general statement to the specific case, or vice versa.”[58]

         There may well be a trend by informal logicians today to incorporate all these fallacies under the rubric of the fallacy of neglecting qualifications.
         
         
         

   4. Unresolved Difficulty in the Fallacy of Accident

      The crux of the identification of the accident fallacy is the difficulty of recognizing the conditions under which an instance of an unconditioned soft generalization is exceptional.
      
      
      
      1. F.C.S. Schiller raises an important issue with respect to the presence of ambiguity when the fallacy of accident is viewed from the point of view of its current definition:

         “[T]he Fallacy of Accident … is defined as an attempt to argue falsely from a rule which is true in general to a case in which, owing to particular circumstances, the rule does not hold. … [T]here always are ‘particular circumstances,’ that there is a real difficulty in applying a rule and arguing from a ‘law’ to a ‘case,’ because it is possible to misapply rules and necessary to find the right rule for the case.”[59]

         Schiller raises an excellent point which is widely recognized by a number of other logicians.[60]
         
         
         
         1. In the example of capital punishment discussed before concerning the killing of a legally condemned person, i.e. …

            “[A]s if a traitor should argue from the sixth commandment, Thou shalt not kill, to prove that he himself ought not be hanged.”

            … abolitionists who oppose capital punishment on moral grounds conclude that the premise “One should never kill another person,” is an absolute universal generalization, not a soft generalization.
            
            So, under this moral point of view, the non-execution of a condemned traitor not only should be, but is, a proper instance of the generalization that killing a person is morally wrong. Thus, the argument that traitors should not be executed is sound, not fallacious.
            
            
            
         2. Distinguishing between usual and unique circumstances is just as difficult as distinguishing between essential and accident circumstances. Hamlin points out:

            “Unfortunately, it is often difficult to say of a property whether it is ‘essential’ or not, and few people these days would be prepared to go so far as to maintain generally that the essential properties of every kind of thing can be uniquely specified. … [I]t is not at all clear how someone who claims that given property is not an essential one will be able to argue against someone else who claims that it is.[61]

            For this reason, among others, the fallacy of accident is most certainly best understood at this time as the fallacy of neglecting qualifications or secundum quid.
            
            
            
      2. Given that the fallacy of accident occurs whenever the conclusion is an unusual, unique, or exceptional case of a soft generalization, how are we to distinguish unusual, unique, or exceptional qualifications from usual, ordinary, or unexceptional qualifications? Consider these three examples of the accident fallacy from a 19th century textbook:

         “[T]he fallacy … which argues from something generally true and undeniable to the same when some special condition is introduced, is also a very frequent and often a very pernicious one. The teetotaller who refuses to give wine to the sick, even when the doctor orders it, on the ground that it is dangerous …”
         
         “[T]he parent who will not correct his pilfering child on the plea that it is cruel to beat children …”
         
         “[T] theologian who condemns Abraham's intention to sacrifice Isaac, on the ground that murder is always unjustifiable …”[62]

      Can we justifiably argue that giving wine to a sick person, beating a child for for minor theft, or condemning someone for intending to murder his son are unexceptional or usual examples of “pernicious” conditions? In cases such as these, changing cultural conditions ought not be thought to generally effect logical judgment.
      
      
      
   5. For the fallacy of accident or the fallacy of the neglect of qualifications to be a useful tool in the identification of fallacious arguments, a reliable criterion of qualification or unique instance need be specified.
      
      
      

5. Historical Frequency of Use of the Fallacy of Accident and Accidentis Fallacia 1800-2019 in Google books Ngram Viewer

   
   
   1. From a historical point of view, the fallacy of accident prior to 1800 was included in either one form of the medieval Scholastic's fallacy of fallacia accidentis (as the confusion between essential and accidental properties) or in one form of the fallacy secundum quid (i.e. the neglect of necessary qualifications).
      
      
      
   2. As can be seen the following Google Ngram graph, the name “fallacy of accident” has superseded one of its designations “fallacia accidentis” in everyday discourse.
      
      

postscript

“Accident is not so much a fallacy in the sense of a common error of reasoning. It could perhaps better be described as a paradox or series of puzzles, a general category for a family of quite subtle types of errors or problems of reasoning. Whatever acci­dent is, it is far from clear that it is the kind of common error of reasoning or fallacy that belongs in the standard treatment of fallacies in logic textbooks (especially textbooks meant for general reading by nonspecialists and introductory students).”

Douglas Walton, Argumentation Schemes for Presumptive Reasoning (1998 Oxford: Routledge, 2013), 148.
doi: 10.4324/9780203811160

”Trying to explain the historical origins of unfamiliar and puzzling terms like ‘accident’ … seems a hopeless and unrewarding task, not to mention the variety of Latin terms and phrases peppered through the traditional textbook treatments. While ignoring exceptions to a rule or qualifications to a generalization did seem important failures to warn students about, nevertheless the complications and puzzles inherent in the standard treatment suggested that it was prudent to bypass this area, restricting coverage to a brief mention of the basic fault, in simple terms.”

Douglas N. Walton, “Ignoring Qualifications (Secundum Quid) as a Subfallacy of Hasty Generalization,” Logique & Analyse no. 129-130 (1990), 114.

Links to Additional Examples and Online Practice Quizzes with Suggested Solutions for the Fallacy of Accident

Test your understanding of the fallacy of accident occurring in the following quizzes:

Fallacy of Accident Examples Exercise
Fallacies of Relevance I
Fallacies of Relevance II
Fallacies of Relevance III