Fragment

style-guide

Elements of Style — Scientific

agent starticle

The Elements of Style, Adapted for Scientific Writing and Journalism

After William Strunk Jr. (1918), adapted to the disciplines of the laboratory and the newsroom.

I. Introductory
III. Elementary Principles Of Composition

I. Introductory

This handbook summarizes the essentials of clear prose as they apply to scientific writing and journalism. The scientist and the reporter face the same fundamental problem: they must convey what is true, precisely and without ornament, to readers who will act on what they read. In the laboratory the consequence of vague writing is failed replication; in the newsroom it is public misunderstanding. The rules that follow address the errors most often committed in both disciplines. Master them here, then study the best practitioners—Darwin’s prose, Faraday’s lectures, the dispatches of Martha Gellhorn, the investigations of John Hersey—for the finer points of style.

III. Elementary Principles Of Composition

Rule 8. Make the paragraph the unit of composition: one paragraph to each topic.

If the subject on which you are writing is of slight extent, or if you intend to treat it very briefly, there may be no need of subdividing it into topics. Thus a brief description of an experimental result, a summary of a journal article, a short account of a single incident in the field, a statement of one finding—any of these is best written in a single paragraph. After the paragraph has been written, examine it to see whether subdivision will not improve it.

Ordinarily, however, a subject requires subdivision into topics, each of which should be made the subject of a paragraph. The object of treating each topic in a paragraph by itself is, of course, to aid the reader. The beginning of each paragraph is a signal to him that a new step in the development of the argument has been reached.

The extent of subdivision will vary with the length of the composition. For example, a short notice of a published study might consist of a single paragraph. One slightly longer might consist of two paragraphs:

A. Summary of the findings.
B. Assessment of their significance.

A research article might consist of paragraphs grouped under the following heads:

A. Statement of the problem and its context.
B. Description of methods and materials.
C. Presentation of results.
D. Interpretation of results.
E. Relation to prior work.
F. Limitations and qualifications.
G. Implications for future inquiry.

The contents of paragraphs C and D would vary with the study. Usually, paragraph C would present the data in logical sequence, beginning with the most central finding. Paragraph D would explain what the results mean and how they bear on the hypothesis stated in A.

A news story might be discussed under the heads:

A. The lead: who, what, when, where.
B. The immediate context and significance.
C. Background and prior developments.
D. Reaction and consequences.

An investigative report might be organized thus:

A. The central revelation.
B. The evidence supporting it.
C. The methods by which the evidence was obtained.
D. Responses from those implicated.
E. The broader pattern or systemic failure.

In treating either of these last two subjects, the writer would probably find it necessary to subdivide one or more of the topics here given.

As a rule, single sentences should not be written or printed as paragraphs. An exception may be made of sentences of transition, indicating the relation between the parts of an exposition or argument. In journalism, short paragraphs are conventional and often necessary for the narrow newspaper column; but even here, the paragraph should contain a complete thought, however compressed.

In quotation, each speaker’s words, even if only a single sentence, form a paragraph by themselves; that is, a new paragraph begins with each change of speaker. This rule applies equally to the interview and to the reported dialogue of witnesses.

Rule 9. As a rule, begin each paragraph with a topic sentence, end it in conformity with the beginning.

Again, the object is to aid the reader. The practice here recommended enables him to discover the purpose of each paragraph as he begins to read it, and to retain this purpose in mind as he ends it. For this reason, the most generally useful kind of paragraph, particularly in exposition and argument, is that in which

(a) the topic sentence comes at or near the beginning;

(b) the succeeding sentences explain or establish or develop the statement made in the topic sentence; and

Ending with a digression, or with an unimportant detail, is particularly to be avoided.

This principle is nowhere more important than in scientific writing, where the reader is often scanning for specific information. A paragraph in a results section that begins with a description of the apparatus and ends, almost incidentally, with the key finding, has buried its purpose. The finding belongs at the top; the apparatus, if it must be mentioned again, supports it.

In journalism the topic sentence is the lead. Every news paragraph earns its place by answering a question the reader has formed after reading the paragraph before it. If a paragraph about a chemical spill begins with the mayor’s response and ends with the volume of the spill, the paragraph has drifted from politics to chemistry; it should be two paragraphs, or the order should be reversed.

If the paragraph forms part of a larger composition, its relation to what precedes, or its function as a part of the whole, may need to be expressed. This can sometimes be done by a mere word or phrase (however; by contrast; these results suggest) in the topic sentence. Sometimes, however, it is expedient to precede the topic sentence by one or more sentences of introduction or transition. If more than one such sentence is required, it is generally better to set apart the transitional sentences as a separate paragraph.

Rule 10. Use the active voice.

The active voice is usually more direct and vigorous than the passive:

We measured the oxygen concentration at hourly intervals.

This is much better than

The oxygen concentration was measured at hourly intervals.

The latter sentence is less direct, less bold, and less concise. If the writer tries to make it more concise by omitting the agent entirely,

The oxygen concentration was measured,

it becomes indefinite: is it the authors, or some earlier researcher, or a technician, that performed the measurement?

The convention in much scientific writing has been to prefer the passive, as if the absence of we conferred objectivity. It does not. The passive merely obscures who did what, and when responsibility must be assigned—who designed the experiment, who administered the drug, who made the error—the passive becomes not merely weak but dangerous. Modern journals increasingly require the active voice for this reason.

This rule does not, of course, mean that the writer should entirely discard the passive voice, which is frequently convenient and sometimes necessary.

The samples were incubated at 37 degrees Celsius for 24 hours.

We incubated the samples at 37 degrees Celsius for 24 hours.

The first would be the right form when the conditions of the experiment are the point; the second, when the actions of the investigators are the point. The need of making a particular word the subject of the sentence will often, as in these examples, determine which voice is to be used.

As a rule, avoid making one passive depend directly upon another.

Original	Revision
The drug was expected to be absorbed within two hours.	We expected the drug to be absorbed within two hours.
The policy was reported to have been abandoned by officials.	Officials reportedly abandoned the policy.

In both the examples above, before correction, the word properly related to the second passive is made the subject of the first.

A common fault is to use as the subject of a passive construction a noun which expresses the entire action, leaving to the verb no function beyond that of completing the sentence.

Original	Revision
An analysis of the blood samples was performed.	We analyzed the blood samples.
An investigation of the allegations was carried out by the committee.	The committee investigated the allegations.
A review of the literature was conducted.	We reviewed the literature.

The habitual use of the active voice makes for forcible writing. This is true not only in narrative principally concerned with action, but in writing of any kind. Many a tame sentence of description or exposition can be made lively and emphatic by substituting a verb in the active voice for some such perfunctory expression as there is, or was observed.

Original	Revision
There was a significant increase in mortality among the exposed group.	Mortality rose sharply among the exposed group.
A correlation was observed between income and educational attainment.	Income and educational attainment correlated strongly.
It was found by the reporters that the fund had been diverted.	The reporters found that officials had diverted the fund.
The decision was made to evacuate the building.	The fire chief ordered an evacuation.

Rule 11. Put statements in positive form.

Make definite assertions. Avoid tame, colorless, hesitating, non-committal language. Use the word not as a means of denial or in antithesis, never as a means of evasion.

Original	Revision
The treatment did not produce a significant effect.	The treatment produced no significant effect.
The mayor did not deny the allegations.	The mayor declined to address the allegations (The mayor sidestepped the allegations).
The results were not inconsistent with the hypothesis.	The results supported the hypothesis.

The last example, before correction, is a double negative that sounds cautious but says nothing the writer is willing to stand behind. If the results support the hypothesis, say so. If the support is weak, say that: “The results offered weak support for the hypothesis.” Hedging through negation is not caution; it is evasion.

All three examples show the weakness inherent in the word not. Consciously or unconsciously, the reader is dissatisfied with being told only what is not; he wishes to be told what is. Hence, as a rule, it is better to express even a negative in positive form.

Original	Revision
not significant	negligible
not many	few
did not succeed	failed
did not consider	ignored (overlooked)
not the same	different
not possible	impossible

The antithesis of negative and positive is strong:

Not correlation, but causation.

Not the disease itself, but the treatment killed him.

Negative words other than not are usually strong:

No vaccine has ever been developed so rapidly.

The journalist who writes “The senator was not entirely forthcoming” means that the senator lied. He should say so.

Rule 12. Use definite, specific, concrete language.

Prefer the specific to the general, the definite to the vague, the concrete to the abstract.

Original	Revision
A number of participants experienced adverse reactions.	Twelve of the forty participants vomited within an hour.
The city has a significant pollution problem.	On nine days last month, ozone in downtown Los Angeles exceeded federal limits.
The experiment yielded promising results.	The treated mice survived an average of fourteen days longer than the controls.

If those who have studied the art of writing are in accord on any one point, it is on this, that the surest method of arousing and holding the attention of the reader is by being specific, definite, and concrete. This principle, which Strunk illustrated with Homer and Browning, applies with even greater force in the disciplines of fact. The scientist who writes “the sample turned color” has wasted an observation; the scientist who writes “the sample turned from pale yellow to deep violet” has recorded one. The reporter who writes “a large crowd gathered” has told us nothing we can verify; the reporter who writes “roughly three hundred people filled the church basement” has told us something we can see.

Consider the difference between these two accounts of the same event:

Original	Revision
The earthquake caused significant damage to structures throughout the region and resulted in a substantial number of casualties.	The earthquake buckled the coast highway in three places, collapsed a parking structure in Northridge, and killed sixty-one people.

The first version could describe any earthquake anywhere. The second describes this earthquake, this damage, these dead. Journalism that relies on words like significant, substantial, major, and key is journalism that has not yet done its reporting. These words are placeholders for facts the writer has not obtained.

In scientific writing the same evasion takes the form of considerable, notable, remarkable, and the particularly empty interesting. A finding described as “interesting” has not been described at all. Say what makes it interesting: does it contradict the prevailing model? Does it replicate a result that had been questioned? The reader should not have to guess.

“This superiority of specific expressions is clearly due to the effort required to translate words into thoughts. As we do not think in generals, but in particulars—as whenever any class of things is referred to, we represent it to ourselves by calling to mind individual members of it, it follows that when an abstract word is used, the hearer or reader has to choose, from his stock of images, one or more by which he may figure to himself the genus mentioned.”

Herbert Spencer’s principle, quoted by Strunk, applies with special force in the sciences, where imprecision is not merely a stylistic fault but a methodological one. Consider:

Original	Revision
A number of factors contributed to the failure of the dam.	Three factors contributed to the failure of the dam: inadequate drainage, substandard concrete, and the removal of a retaining wall during construction.

Rule 13. Omit needless words.

Vigorous writing is concise. A sentence should contain no unnecessary words, a paragraph no unnecessary sentences, for the same reason that a drawing should have no unnecessary lines and a machine no unnecessary parts. This requires not that the writer make all his sentences short, or that he avoid all detail and treat his subjects only in outline, but that he make every word tell.

Many expressions in common use violate this principle. Scientific and journalistic prose are especially afflicted:

Original	Revision
at this point in time	now
in the event that	if
a sufficient number of	enough
has the ability to	can
on a daily basis	daily
in order to determine	to determine
it is interesting to note that	(omit entirely, or state what is interesting)
prior to the commencement of	before

In especial the expression the fact that should be revised out of every sentence in which it occurs.

Original	Revision
owing to the fact that	since (because)
despite the fact that	although
draw attention to the fact that	note (point out)
in view of the fact that	since (because)
the fact that the vaccine was effective	the vaccine’s effectiveness
the fact that the bridge collapsed	the collapse of the bridge

Which is, who was, and the like are often superfluous.

Original	Revision
Dr. Salk, who was the inventor of the polio vaccine	Dr. Salk, inventor of the polio vaccine
Chernobyl, which was the site of the worst nuclear disaster in history	Chernobyl, site of the worst nuclear disaster in history

As positive statement is more concise than negative, and the active voice more concise than the passive, many of the examples given under Rules 11 and 10 illustrate this rule as well.

A common violation of conciseness in scientific writing is the presentation of a single complex finding, step by step, in a series of sentences which might to advantage be combined into one.

Original	Revision
The patients were divided into two groups. One group received the drug. The other group received a placebo. The study was double-blind. Neither the patients nor the researchers knew which group received the drug. The study lasted twelve weeks. (39 words.)	In a twelve-week double-blind trial, patients received either the drug or a placebo. (13 words.)
A fire broke out at the refinery on Tuesday. The fire spread to three storage tanks. The fire burned for two days. Five workers were injured. Two of them were hospitalized. (31 words.)	A fire at the refinery on Tuesday spread to three storage tanks and burned for two days, injuring five workers, two of whom were hospitalized. (25 words.)

In journalism, wordiness does not merely bore the reader; it buries the news. The reporter who takes thirty-one words to say what can be said in twenty-five has not merely wasted six words; he has delayed the reader from reaching the next fact by the time it takes to read them.

Rule 14. Avoid a succession of loose sentences.

This rule refers especially to loose sentences of a particular type, those consisting of two co-ordinate clauses, the second introduced by a conjunction or relative. Although single sentences of this type may be unexceptionable, a series soon becomes monotonous and tedious.

An unskilful writer will sometimes construct a whole paragraph of sentences of this kind, using as connectives and, but, so, and less frequently, who, which, when, where, and while. Observe:

The study was conducted over a period of six months, and it involved more than a thousand participants. Dr. Hoffman led the research team, and the university provided the funding. The early results were encouraging, but the final analysis revealed several confounding variables. The team published its findings in the Lancet, and the paper attracted considerable attention. A follow-up study was planned, which was expected to begin the following year.

Apart from its emptiness, the paragraph above is weak because of the structure of its sentences, with their mechanical symmetry and sing-song. The reader feels he is being walked through a series of rooms, each of the same shape, connected by the same kind of door. Nothing in the sentence structure tells him which facts matter most.

If the writer finds that he has written a series of sentences of the type described, he should recast enough of them to remove the monotony, replacing them by simple sentences, by sentences of two clauses joined by a semicolon, by periodic sentences of two clauses, by sentences, loose or periodic, of three clauses—whichever best represent the real relations of the thought. The good journalist varies sentence structure not for variety’s sake but to signal importance: the short declarative sentence for the central fact, the longer construction for context and qualification.

Rule 15. Express co-ordinate ideas in similar form.

This principle, that of parallel construction, requires that expressions of similar content and function should be outwardly similar. The likeness of form enables the reader to recognize more readily the likeness of content and function.

The unskillful writer often violates this principle, from a mistaken belief that he should constantly vary the form of his expressions. It is true that in repeating a statement in order to emphasize it he may have need to vary its form. But apart from this, he should follow the principle of parallel construction.

Original	Revision
The vaccine reduced infection rates, and hospitalizations were also seen to decline.	The vaccine reduced infection rates and lowered hospitalizations.
Formerly, the data was collected by hand, while now computerized methods are employed.	Formerly, the data was collected by hand; now it is collected by computer.

The left-hand version gives the impression that the writer is undecided or timid; he seems unable or afraid to choose one form of expression and hold to it. The right-hand version shows that the writer has at least made his choice and abided by it.

By this principle, an article or a preposition applying to all the members of a series must either be used only before the first term or else be repeated before each term.

Original	Revision
The study examined effects on the liver, the kidneys, cardiovascular system, and lungs.	The study examined effects on the liver, the kidneys, the cardiovascular system, and the lungs.
In rats, mice, or in primates	In rats, mice, or primates (In rats, in mice, or in primates)

Correlative expressions (both, and; not, but; not only, but also; either, or; first, second, third; and the like) should be followed by the same grammatical construction. Many violations of this rule arise from faulty arrangement; others from the use of unlike constructions.

Original	Revision
The drug was both effective in reducing symptoms and it lowered mortality.	The drug both reduced symptoms and lowered mortality.
Not the data, but how it is interpreted.	Not the data, but the interpretation.
Either the government must regulate emissions or face the consequences of inaction.	The government must either regulate emissions or face the consequences of inaction.
The report concluded, first, the inefficiency of the program; second, that it was poorly managed.	The report concluded, first, that the program was inefficient; second, that it was poorly managed.

Parallel construction is especially valuable in scientific writing, where the reader must compare results across conditions. If one condition is described with a noun phrase and the next with a clause, the asymmetry slows comprehension. The reader who encounters “Group A showed improvement, and a decline was observed in Group B” must work to see that these are co-ordinate findings stated in unlike forms. “Group A improved; Group B declined” lets the comparison land at once.

The position of the words in a sentence is the principal means of showing their relationship. The writer must therefore, so far as possible, bring together the words, and groups of words, that are related in thought, and keep apart those which are not so related.

The subject of a sentence and the principal verb should not, as a rule, be separated by a phrase or clause that can be transferred to the beginning.

Original	Revision
The researchers, after reviewing three years of patient records, concluded that the drug was ineffective.	After reviewing three years of patient records, the researchers concluded that the drug was ineffective.
The senator, in a statement released late Friday, denied any wrongdoing.	In a statement released late Friday, the senator denied any wrongdoing.

The objection is that the interposed phrase or clause needlessly interrupts the natural order of the main clause. Usually, however, this objection does not hold when the order is interrupted only by a relative clause or by an expression in apposition.

The relative pronoun should come, as a rule, immediately after its antecedent.

Original	Revision
The researchers published a study on cardiac arrest in the New England Journal, which enrolled over ten thousand patients.	The researchers published in the New England Journal a study on cardiac arrest that enrolled over ten thousand patients.
The city awarded a contract to Meridian Construction, a firm with no prior government work, which was worth $14 million.	The city awarded Meridian Construction, a firm with no prior government work, a contract worth $14 million.

In scientific prose, the separation of a number from its unit, or a measurement from the thing measured, is a frequent source of confusion:

Original	Revision
The temperature of the solution, which had been heated for three hours, rose to 85 degrees.	The solution, heated for three hours, reached 85 degrees.

Modifiers should come, if possible, next to the word they modify. If several expressions modify the same word, they should be so arranged that no wrong relation is suggested.

Original	Revision
The study only found two cases of adverse reaction.	The study found only two cases of adverse reaction.
Nearly all the samples were not contaminated.	Not all the samples were contaminated (Most of the samples were uncontaminated).
The chief medical officer will give a briefing on Monday morning at the hospital, to which the press is invited, on the status of the outbreak at 10 A.M.	On Monday morning at 10 A.M., the chief medical officer will give a briefing at the hospital on the status of the outbreak. The press is invited.

Rule 17. State your evidence before your claim.

In scientific writing and in investigative journalism, a claim is only as strong as the evidence behind it. If the evidence comes after the claim, the reader must accept the claim on faith while he waits for its support; if it comes before, he arrives at the claim already persuaded.

Original	Revision
The vaccine is safe for children under five. In clinical trials involving six thousand children, no serious adverse events were recorded.	In clinical trials involving six thousand children under five, no serious adverse events were recorded. The vaccine appears safe for this age group.
The company has committed fraud. Documents obtained under subpoena show that executives falsified revenue figures for three consecutive quarters.	Documents obtained under subpoena show that executives falsified revenue figures for three consecutive quarters. The evidence points to systematic fraud.

This principle should not be confused with the journalistic convention of leading with the most important fact. In a news story, the lead may well state the conclusion—“Three executives were indicted Tuesday on charges of fraud”—because the reader needs the news at once. But the paragraphs that follow should supply the evidence in its natural order, so that by the time the reader reaches the end, he has not merely been told what happened but shown why it matters.

In scientific papers the same principle governs the relation between the Results and Discussion sections. Present the data first; interpret it second. The writer who mingles interpretation with results forces the reader to disentangle fact from opinion, a labor that should not be required.

A useful test: if the evidence can be removed without altering the claim, the claim is unsupported. If the claim can be removed without altering the evidence, the claim was probably unnecessary.

Rule 18. Place the emphatic words of a sentence at the end.

The proper place in the sentence for the word, or group of words, which the writer desires to make most prominent is usually the end.

Original	Revision
The outbreak has killed more than three hundred people so far this year in West Africa.	This year in West Africa, the outbreak has killed more than three hundred people.
The compound is primarily used for treating malaria, because of its low toxicity.	Because of its low toxicity, the compound is primarily used for treating malaria.

The word or group of words entitled to this position of prominence is usually the logical predicate, that is, the new element in the sentence, as it is in the second example.

The effectiveness of the periodic sentence arises from the prominence which it gives to the main statement.

After four years of fieldwork in three countries, after the failure of two earlier trials and the withdrawal of their principal funder, after setbacks that would have ended most research programs, the team announced a working vaccine.

In the twelve months since the first case was identified in a village whose name most of the world had never heard, through a chain of transmission that crossed four national borders and overwhelmed the resources of three governments, the virus infected an estimated forty thousand people.

The other prominent position in the sentence is the beginning. Any element in the sentence, other than the subject, may become emphatic when placed first.

Fraud on this scale the regulators had never seen.

So slight a difference in dosage, so vast a difference in outcome.

A subject coming first in its sentence may be emphatic, but hardly by its position alone. In the sentence,

A single whistleblower brought down the company,

the emphasis upon whistleblower arises largely from its meaning and from the context. To receive special emphasis, the subject of a sentence must take the position of the predicate.

From a single contaminated well came the epidemic that killed ten thousand.

The principle that the proper place for what is to be made most prominent is the end applies equally to the words of a sentence, to the sentences of a paragraph, and to the paragraphs of a composition. The final paragraph of a research paper should not trail off into boilerplate about “future work”; it should leave the reader with the finding that matters most. The final paragraph of a news story should not repeat the lead; it should deliver the detail or quotation that gives the story its lasting force.

Rule 19. Distinguish reporting from interpretation.

The reporter reports; the editorialist interprets. The scientist presents data; the theorist explains it. These functions may be performed by the same person, but they must not be performed in the same sentence without clear demarcation. The reader has a right to know which is which.

Original	Revision
The city’s reckless policy of dumping waste in the river has contaminated the water supply.	The city has dumped waste in the river for six years. Three downstream wells now show nitrate levels above the federal limit.
The irresponsible delay in publishing the trial data suggests a cover-up.	The trial ended in March. The data were not published until December. The researchers have offered no explanation for the delay.

In the first pair, the original contains two interpretive words—reckless and contaminated (implying causation)—embedded in what presents itself as a factual statement. The revision separates the fact (dumping) from the fact (nitrate levels) and lets the reader draw the connection. If the writer wishes to assert causation, he should do so explicitly and supply the evidence.

In the second pair, the words irresponsible and cover-up are opinions dressed as news. The revision states only what can be verified: the dates and the silence. The reader may draw his own conclusions; the reporter has given him the facts to draw them from.

In scientific writing, the same discipline governs the distinction between results and discussion. A sentence in the results section should not contain the word because; that word belongs in the discussion, where interpretation is expected and welcome. The writer who observes that “the mice died because the dosage was too high” has, in the results section, made an interpretive claim. He should write: “All mice in the high-dosage group died within 48 hours. In the low-dosage group, none died.” The explanation belongs later.

Rule 20. Let the data speak; do not speak for the data.

The temptation in both science and journalism is to tell the reader what to think before he has been given the material to think with. Resist it. Present the evidence fully, arrange it clearly, and trust the reader to grasp its significance. If the significance is not self-evident, you have not presented the evidence well enough.

Original	Revision
The results were stunning: the treatment group showed a 40% reduction in mortality.	The treatment group showed a 40% reduction in mortality (95% CI: 28–51%, p < 0.001).
The shocking revelation that the agency had been monitoring citizens without warrants rocked the nation.	The agency monitored the phone records of 1.2 million citizens over a period of three years without obtaining warrants.

In the first pair, the word stunning asks the reader to be impressed before he has seen the confidence interval. Give him the numbers; he will decide for himself whether they are stunning. In the second pair, shocking and rocked the nation are the writer’s emotions projected onto the reader. The fact—warrantless surveillance of 1.2 million people for three years—is shocking enough without instruction.

This does not mean that the writer should never interpret or comment. It means that the evidence should always be presented first and in sufficient detail that the interpretation can be judged. The astronomer who says “the data suggest a planet” earns attention because the data have been shown. The astronomer who says “we have discovered a planet” and then shows ambiguous data has squandered his credibility.

A useful test: remove every adjective of magnitude (significant, dramatic, unprecedented, massive) from your draft. Read it again. If the facts no longer sound impressive, the adjectives were doing work the facts should have done. If the facts still sound impressive, the adjectives were superfluous. In neither case were they needed.