Eight Problems of Logic in Writing
Persuasive writing avoids these rhetorical flaws
(An excerpt from Stet
Again!)
Few things ruin a piece of good writing quicker than a
bad argument. The theme may be as clear as distilled water.
The words may fit the topic, and the sentences may flow
like warm honey. The rhetorical flourishes may all be subtle,
and the tone seductive. But if the writer falls into the
many logical errors to which all argument is prone, the
whole effort is jeopardized.
Here we look at eight logic problems in writing: faulty
generalization, false dilemmas, the post hoc ergo propter
hoc argument, the ad hominem argument, question
begging, the improper appeal to authority, imperfect or
false analogies, and the non sequitur. Examples and
editorial strategies are suggested for each.
Faulty generalization
This is perhaps the most common error in argument construction.
Broad generalizations about difficult and controversial
topics seldom stand scrutiny; hence, one useful rule of
thumb for editors is: The stronger the generalization,
the more bells should go off in the editor's mind.
Fortunately, faulty generalizations give off powerful signals.
Often they are preceded by absolute words like all,
always, anyone, every(one)(thing),
only, never, and none. Frequently they
are flagged by superlatives: best, greatest,
most, least. The recommended editorial treatment
is to qualify the absolutes and superlatives.
Example and strategy: "Hamlet is
the best play ever written in English" can be easily changed
to "Many scholars agree that Hamlet is one of the
best plays..." once the author agrees.
False dilemma
This error usually takes the form of presenting two (or
three) alternatives as if they were mutually exclusive,
when they are really just the obvious options or the ones
the writer finds easiest to manipulate. The words and phrases
that commonly signal false dilemmas are these: either...or;
the only alternative/possibility/choice is; since
X is obviously false/has not worked, Y is clearly true/our
only choice. The recommended treatment is to break open
the dilemma by broadening the statement or, where called
for, the entire argument.
Example and strategy: "Either the courts
or the individual conscience is entitled to speak on abortion,
but not both." Here, the editor must point out the logical
problems created by the implication that there are only
two possible camps to be in. It is well within the realm
of logic for both conscience and courts to speak,
for example. Or the dilemma can be broken open by querying
the role of the church and other moral leaders.
Post hoc ergo propter hoc
This error is among the most common fallacies of thinking
and therefore of writing. The Latin phrase means "after
this, therefore on account of this." The error lies in assuming
that establishing a chronological relationship automatically
establishes a causal one. Some of the key signals to look
for, of course, are words that highlight conclusory statements:
therefore, thus, consequently, as
a result, accordingly, on account of.
But not all post hoc errors are that easy to spot.
Example and strategy: "The period following
the U.S. hydrogen bomb tests in the South Pacific was
marked by uncharacteristic and dramatic weather changes,
directly related to the tests." As the statisticians say,
"Correlation is not causality." In the absence of strong
evidence to bolster the writer's implication, the statement
draws a hidden and unwarranted conclusion and therefore
should be qualified, perhaps by inserting some phrase
like "leading many meteorological experts to conclude
that the changes were..." before the word "directly,"
followed by the evidence itself.
Argument ad hominem
This logic error substitutes discrediting one's opponent
for discrediting the opponent's arguments. The fallacy is
sometimes difficult to identify, primarily because the individual
involved may be relevant to the issue. For example, the
1947 conviction of James Michael Curley for mail fraud while
in office became an issue in his 1949 campaign for reelection
as mayor of Boston. But in most cases, the ad hominem
argument should not be allowed to stand.
Example and strategy: "Karl Marx's poor
performance as a provider for his own family is reason
enough to doubt the validity of his social and economic
theories." Marx's ability to earn a living may or may
not have anything to do with the validity of his theories,
which are quite large in scale. The editor must ask the
author either to present historical and economic facts
discrediting the theories, independent of Marx's performance
as a provider, or show that Marx's application of his
theories in his own household caused his family economic
hardship.
Question begging
Question begging comes in two main forms. One is a form
of circular or tautological reasoning. Setting out to make
a point, the writer posits the conclusion the reader is
expected to make, then supports or "proves" it by some restatement
or assertion of the original point. The conclusion gives
rise to the reason, which supports the conclusion.
Example and strategy: "The study of history
gives excellent insights into human nature because one
can learn much about personal character by learning about
the events and circumstances of the past." Here, the reason
provided for the proposition in the first half of the
sentence is simply the proposition's restatement in slightly
different words in the second half. The editor needs to
flag the error with "Tautol." or "Circ. Arg.,"
improve the sentence simply by breaking off the last half
of it, and follow up by asking the author to give a concrete
example.
A second form of question begging is the smuggled assumption.
In this case the writer hides an assumption in the statement
of a proposition and proceeds to prove or support the proposition
by drawing on the hidden assumption.
Example and strategy: "Capital punishment
is forbidden in Sweden and its homicide rate is lower
than ours. We should therefore abolish capital punishment."
The hidden assumption is that the relationship between
the absence of capital punishment and Sweden's low homicide
rate is a causal one. What should be proven — that we
should abolish capital punishment — is supported by this
assumption. A useful editorial strategy is to state the
question that is, in fact, begged by the offending sentence:
"Does lack of capital punishment cause low murder
rate? Are there other factors?"
Improper appeals to authority
Appeals to authority are common in all writing: "As Churchill
once said...," "Four out of five doctors agree...," "According
to the AMA Journal...," "Research shows...." But
appeals to authority have different weights, and not all
authorities are trustworthy beyond their fields of expertise.
Henry Kissinger's statements about Russian cuisine are less
authoritative than his statements about Soviet foreign policy.
Nor is the evidence provided by polls necessarily authoritative.
Example and strategy: "Forty-seven percent
of New York viewers characterized William Hurt's performance
in The Accidental Tourist as his best ever." The
editor could pencil in any or all of the following questions
in the margin: "What New York viewers?" "What
are their qualifications for judging acting?" "What
was the size of the sample?" Other questions editors
can put to appeals to authority include: "Is this authority
competent in this field?" "Is the authority providing
evidence, reasoned analysis, or merely opinion to support
the argument?" "Is the authority known to have
a specific bias?" "Is the authority's statement
up-to-date?"
Imperfect or false analogy
Reasoning by analogy attempts to make a case by comparing
something known with something not known or not as well
known. The potential for error lies in an unwarranted leap
of logic: The writer jumps from the fact that two partners
in the analogy share one or more characteristics to the
assumption that the second partner in the analogy shares
an additional characteristic of the first, which
characteristic is relevant to the matter at hand. The logical
argument of the false analogy has this syllogistic form:
X and Y share characteristics a, b, and
c.
X has the further characteristic z.
Therefore, Y must share characteristic z.
Example and strategy: "Education cannot
possibly prepare a couple for parenthood. Trying to educate
them for such a task is like trying to teach them to swim
without putting them in the water; it can't be done."
Here the argument assumes a set of shared characteristics
between parenthood and swimming — overcoming fears, learning
new skills, adapting to a new environment, adopting new
behavior patterns, and so forth. But, the editor can point
out, swimming and parenthood are also very different from
each other. A terse "Analogy doesn't work. Swimming
skills physical; parenting skills psychosocial" should
wake the author up.
The great value of analogical argument lies in the insight
revealed by the unexpected juxtaposition of two elements.
But all analogies break down when pressed too far. The trick
the editor must learn is to find where the line is and not
let the writer cross it.
Non sequiturs
A non sequitur is a conclusion or inference that
"does not follow" from the premises offered. Some examples
of non sequiturs are
Delton drinks too much. He probably beats his kids, too.
Phogbound is a politician. Can we trust him?
Elyse has perfect Sunday school attendance; I'll bet
she's well behaved at home.
Non sequiturs are similar to false analogies in
that both depend on the transfer of imputed characteristics
from one context to another to make the point. But non
sequiturs are more fallacious forms of reasoning than
false analogies because the relationships on which they
depend are usually more tenuous. There may be a certain
"reasonableness" to non sequiturs; for example, Elyse's
perfect Sunday school attendance may lead us to think she's
a "nice little girl," but there may be another, better reason
that explains her behavior (perhaps her parents make her
go); she may be a perfect hellion at home. The editor can
simply query: "Perfect attendance = good behavior @ home?"
Logic is crucial to sound argument and persuasion, and
thus to clear writing. It is, as Jean de la Bruyère
has said, "the art of making truth prevail." But logic can
also be overdone. Some of our most profound experiences
— love of country, romantic passion, religious awe, aesthetic
joy — defy logic, and these too must be expressed. The
beginning editor will thus slowly learn the lesson taught
by Pascal, which experienced editors already know, that
"the last function of reason is to know the infinity of
things that surpass it."
—Bruce O. Boston
Privacy Policy |