Educational Workshop

EW13: Scientific writing: how to go about it

Convenors: Judith Crane (Paris, FR) Panayotis T. Tassios (Athens, GR)

Faculty: Sherwood Gorbach (Boston, US) Trish Groves (London, UK) George Schmid (Geneva, CH) Panayotis T. Tassios (Athens, GR)

INTRODUCTION

Scientific writing: how to go about it

Respect the acknowledged Include the features conventions of a research paper that could make your paper stand out among the many

A topic worthy of publication… … and a style that makes this obvious

An honest assessment of the scope… … and a text that impresses without exaggerating

The right group of authors… … and a coherent text despite multiple authors

A Title that is specific to the content… … and also appeals to the non-specialist

An Abstract that is clear… … and also gets attention

An Introduction that poses a question… … and keeps the reader reading

Methods and Results that are complete… … and as concise as possible

Data that are accurate… … and that make a (citeable) point

A Discussion that puts ideas into a context… … and also generates a novel idea

Careful use of language and terminology… … and one unique statement

A Conclusion that is relevant … and also memorable

Judith Crane, Managing Editor, Clinical Microbiology and Infection ([email protected]) http://mc.manuscriptcentral.com/cmi

Gorbach –What an editor wants to see

WHAT AN EDITOR WANTS TO SEE IN A SUBMISSION

Sherwood L. Gorbach, MD Editor, Clinical Infectious Diseases Tufts University School of Medicine Boston, MA

EDITOR’S DESIDERATA

Original, informative, accurate and n topical Newsworthy n Generalizable beyond confines of n authors’ locale. Broad applicability Discourage case reports, unless new n information or novel treatment

PUBLISHING AN ARTICLE IN CLINICAL INFECTIOUS DISEASES

Write a clear abstract with concise n conclusions. Include a strong cover letter with the n rationale for publishing this paper. Use clear English; avoid misspellings and n grammatical errors. Must not be BORING! n Gorbach –What an editor wants to see

CID POLICIES

An article must be directly related to clinical issues n –purely laboratory or animal studies are declined. Original clinical research papers are preferred over n case reports. Invited mini-reviews in 16 topic areas are solicited n by Special Editors. Some reviews are author- initiated. Editorial commentaries are invited for ~4 articles n per issue.

THE EDITOR RELIES ON REVIEWERS: What Makes a Good Reviewer

The characteristics of reviewers i.e. those actively n involved in research, occupying academic positions, in clinical practice, and members of research funding bodies, have little association with the quality of the reviews they produced. People nearer 40 than 60 years of age produced n better reviews. Review quality increased with time spent on a n review, up to 3 hours but not beyond.

Black et al., JAMA, 1998

A REVIEWER’S MANDATE

Help the editor n Help the author n Help Mankind! n The review should be n Comprehensive n Critical n Even-handed n Adapted from M. Callaham, D. Schriger, R.J. Cooper, 2004-5 Gorbach –What an editor wants to see

CHALLENGES TO ACADEMIC ID JOURNALS - 2009

Number of subscribers is limited, especially in a n small specialty like ID Specialty journals must publish articles on topics n with limited general interest but with high importance to specialists, e.g., Buruli Ulcer, Rift Valley Fever, treatment of Leishmaniasis. This policy lowers the IF since there are few citations n Specialty journals have lower IF, which means n Less interest in publishing articles by academics n Lower prestige n Fewer readers n

CHALLENGES TO ACADEMIC ID JOURNALS Special issues associated with a “Society” journal Editor reports to a Publications Committee, which is often n composed of elected members who have little experience in academic publishing. Famous examples of firing an editor for insubordination and refusal to follow demands of the Society (e.g., JAMA, NEJM). Profits of the journal, if they exist, go to support the Society, n which may produce pressures to increase advertising, sales, supplements. In commercial publishing it’s dog eat dog, and in academic n publishing, it’s said to be the reverse. Society members have a sense of entitlement about n publishing their own papers. Editor must devote pages to Society affairs: news, committee n reports, Guidelines –usually without peer review.

WHAT MAKES A JOURNAL EDITOR LOSE SLEEP

Lyme Disease IDSA Guidelines published in CID (2007) n evinced: multiple angry letters to the editor n threats to the authors n lawsuits against IDSA n a criminal investigation by the Attorney General of the State of n Connecticut (restraint of trade). From In the Literature, CID, Mar 15, 2008: “The report of n viscerotrophic leishmaniasis in 8 US soldiers who had been deployed to eastern Saudi Arabia during the first Gulf War served as a warning that soldiers would likely be returning from the latest military misadventure with similar problems.” Two letters were received from medical officers in the US Army n Medical Corps protesting the mixing of “pejorative” political asides into a scientific report. Gorbach –What an editor wants to see

FATE OF ARTICLES REJECTED BY A GENERAL MEDICAL JOURNAL

350 articles rejected by Ann Int Med n (1993-4) 240 (69%) published subsequently in a n specialty journal; mean time to appear 552 days (found in PubMed) Mean IF 3.09 vs. 9.6 (Ann Int Med) n

Cited by Rowe, RC. "Publish or Perish." Drug Discovery Today. 9:590-591, 2004 Groves –Impression on reviewers

How to make the most favourable impression on reviewers

Dr Trish Groves Deputy editor, BMJ [email protected]

What I plan to cover

Peer review

• how reliable is it anyway?

• models at different journals

• how to please reviewers

How reliable is peer review? Groves –Impression on reviewers

“Peer review is like democracy, which is, to use Churchill's phrase, the worst form of government except all those other forms that have been tried from time to time.”

Rennie D. More Peering into Editorial Peer Review. JAMA 1993;270:2856-58.

Is peer review a game of chance?

Neff and Olden (2006) used a Bayesian approach and citation data from biological journals to model the process and found that –the process includes a strong “lottery” component, independent of editor and reviewer integrity –top journals tend to use prescreening then three reviewers; about a quarter of published papers still poorly cited –element of chance is greater if journals use only two referees and do no prescreening (or if only one editor prescreens); about half of published papers poorly cited

How often do two reviewers agree?

At journals eg NEJM (Ingelfinger F 1974) • rates of agreement “moderately better than chance” • agreement greater for rejection than acceptance

At funding bodies • Cole et al, 1981. Real vs sham panel, 75% agreement • Hodgson C, 1997. Two real panels reviewed the same grants, with 73% agreement

Are two reviewers enough? • Fletcher and Fletcher 1999. At least six reviewers, all favouring rejection or acceptance, are needed to yield a reliable conclusion Groves –Impression on reviewers

Should we mind if reviewers don’t agree?

No, we shouldn’t, if: • reviewers are chosen for differing views • very high reliability might mean that all reviewers think the same • we accept that false positives and false negatives are inevitable • reviewers often advise rather than decide

But…

Biases in peer review

Author-related • prestige (author/institution) • gender • where they live and work Paper-related • positive results • English language Reviewer-related • competing interests • personal issues

How to minimise bias

Closed Open • declared competing • declared competing interests interests • double blind review • open (signed) review • single blind review • open (to all) review –authors’ identity • post-publication masked openness –reviewers’ identity masked Groves –Impression on reviewers

Peer review at different journals

Which models at which journals?

• named, pre-publication review: BMJ, JRSM, BioMed Central medical journals • open and permissive peer review (preprints published if authors successfully solicit 3 reviews): Biology Direct • community peer review: Atmospheric Chemistry and Physics, Nature trial • two step review (invited signed review then community review): Journal of Interactive Media in Education • permissive peer review, post-publication commentary and annotation: PLoS One • no formal peer review, post-publication commentary: Nature Proceedings

Matt Hodgkinson http://journalology.blogspot.com/2007/06/open- peer-review-community-peer-review.html Groves –Impression on reviewers

BMJ open (signed) peer review What we say to reviewers

“The BMJ uses open peer review so that authors know who has reviewed their work. This means that you will be asked to give your name and position, and any relevant competing interests, in your report on any article we send you. It does not mean that authors should contact you directly…nor should you contact the author directly.

… We will pass on your signed report to the author, so please do not make any comments that you do not wish the author to see.

If you experience any adverse event arising from open peer review, or would like to tell us your views, please email [email protected]”

How to please reviewers

Before you start your study

• have a clear research question • seek statistical advice • use the right study design • keep an open mind and minimise bias • agree who will be principal investigator • agree who will be authors and contributors • agree to publish even negative results Groves –Impression on reviewers

Which journal should you choose?

Points to consider:

• audience • prestige • impact –more than just Impact Factor –will people read your work? Respond to it? Debate it? Replicate it? Alert others to it? –will anyone use your findings? –will your work reach the public?

How to please editors and reviewers: submission

• choose the right journal • read the journal • follow advice to authors • submit online and follow instructions • suggest reviewers if journals invites you to

What do editors and reviewers want?

• importance • originality • relevance to the audience • real potential to improve decision making • truth and transparency • clear writing that people want to read • excitement/ “wow” factor Groves –Impression on reviewers

Clear writing

• International Committee of Medical Journal Editors uniform requirements for manuscripts submitted to biomedical journals http://www.icmje.org/

• reporting guidelines for research, at the EQUATOR network resource centre http://www.equator-network.org/

Transparency and integrity

BMJ’s transparency policy will help you to report work honestly and fully

http://resources.bmj.com/bmj/authors/editori al-policies/transparency-policy Groves –Impression on reviewers

How to please editors and peer reviewers: communication

• make sure the message is clear in the paper and abstract, not just in the cover letter • avoid salami • include relevant extras eg STROBE checklist • cite (and send) any closely related papers • send previous peer review reports • communicate clearly and promptly

Revisions

Provide:

• revised manuscript • all necessary supporting information • detailed covering letter explaining how you have responded (or not) to comments and other guidance

Appeals

• BMJ always willing to consider first appeals (letter first; not revised paper initially) • more success if authors respond in detail to editors’ and reviewers’ criticisms • perhaps 20% accepted on appeal • no second appeals Groves –Impression on reviewers

Schmid –Logical presentation

Logical Thinking, Logical Presentation

How to present your data in the best light

George Schmid, M.D., M.Sc. Supplements Editor, CM I Department of HIV/AIDS World Health Organization Geneva, Switzerland [email protected]

August Almaty, Kazakhstan [email protected]

Outline

1. What (type of) journal? 2. What type of article? 3. What type of writing?

What (type of) Journal?

Not easy! Schmid –Logical presentation

1. What (type of) Journal?--considerations

•The content of your article •The audience you want to reach •The speed with which you want to reach the audience •The prestige of the journal •Can you "afford" a rejection?

2. What Type of Article?--considerations

•Every journal has differing types of articles •Choose the one that most suits your work •Brevity

3. What Type of Writing? Schmid –Logical presentation

Think of your paper as a symphony orchestra

There are many parts, and the contribution of each must come together appropriately and harmoniously to achieve a successful presentation of the what the composer intended

A Typical Paper

1. Title page 2. Abstract 3. Introduction 4. Materials and Methods 5. Results 6. Discussion 7. Acknowledgements 8. Tables and Figures 9. References

1. Title Page

• Title - Specific - Not technical - Think of audience - "Catchy" is okay, but not "cute" • Authors - First author— person who did most of the work (and submitting for publication takes additional work) - Last author, the "senior author"— often the mentor or overall guide to the paper - Everyone else • Must have contributed (see guidance) • First three authors are most visible because of citation • Include word count for abstract and text Schmid –Logical presentation

2. Abstract (summary)

•Includes: - Purpose - Methods - Results - Conclusions - Critical mass •Less technical than paper (think of audience) •No abbreviations •Concise

3. Introduction

•Summarize relevant literature •What is unknown? •Why did you ask your question? •What question did you ask?

Must be interesting!

4. Materials and Methods

•How did you answer your question? •Must be enough information to allow another scientist to repeat your experiment, and, fully understand what you did •Diagrams, tables, flowcharts— all okay •No results, although preliminary or previous work that guided your work can be cited1 •Ethics considerations here

1 For example, if you were studying the epidemiology of HIV infection in injection drug users, and preliminary work indicated users came to a park and shared needles only between 18:00- 24:00, you could cite that as why you surveyed users in the park only between those hours Schmid –Logical presentation

5. Results

• Presented in the order that you proceeded with your experiment/study • No need to include all data - "It was interesting… ." often=don't include it • Tables and figures useful - Summarize finding(s) in words and refer to the table/figure - Do not use if findings can be summarized in text • Use numbers and percentages • Be honest! - "The drug cured 33.3% of infected mice, another 33.3% remained infected, and 33.3% (the third mouse) got away."

6. Discussion

• Highlight most important findings but do not simply repeat them • How do your results relate to the question you asked? • Do the results (fully) support your hypothesis? - If not, why not? • Are your results consistent with those of other investigators? • How do your results related to "the big picture" and what work is needed? • End with one-sentence conclusion of your work and why it is relevant

7. Acknowledgements

•Cite those who helped, and why •Don't forget helpful reviewers Schmid –Logical presentation

8. Tables and Figures

•A complex, but fun, area— get good at it •Look through other papers for examples of ways of presenting data or consult a statistician (often good at this) •"A picture is worth a thousand words"

9. References

•Ideally done with EndNote or Reference Manager •Use the correct style and format for the journal you are submitting to, although "Vancouver Convention" is acceptable

Make it Look Good

• A paper that is poorly presented suggests the work in it was done the same way • Ensure the title page looks good and has all information • Number all the pages • Use a running head • Proofread last version before submitting • No misspellings, be consistent in use of abbreviations, ensure spacing/indents are consistent, etc. - If English is not your first language and you are submitting to an English language journal, use an editor (coauthor, colleague, hired, etc) Schmid –Logical presentation

Summary

•Constructing a paper takes time •Constructing a paper should be fun (even the last draft or two) •The result, a paper, is a reflection on the quality of your work, although a balance between perfection and very good must be taken

Do not let the excellent be the enemy of the good

Tassios –English usage

Simple words for complex matters

Panayotis T. Tassios

Associate Editor Clinical Microbiology and Infection Department of Microbiology, Medical School National and Kapodistrian University of Athens, Greece ([email protected])

Outline

• Why write well? • Before beginning • Numbers and words • Short is beautiful (and wise) • Simple words • The right words • Repetition is not redundant • Check for logic • Critical (re)reading • Conclusions

Why write well?

Write quickly, and you will never write well; write well, and you will soon write quickly. –Quintilianus

•The cynical answer: to simplify the publishing process by making it easier for the editor and reviewers to appreciate your results •The obvious answer: to present yourself and your work in the best light •Good results require, and warrant, good writing Tassios –English usage

Before beginning

•Learn from the papers you have enjoyed reading and found important – •Find expressions or styles of writing that: –caught your interest, –contributed to your understanding, –gave you ideas for your own writing •Follow these examples (initially); you will end up providing an example for others

Numbers and words

•Numbers are also words •Tables and figures should tell the entire story •Therefore: –Label figures comprehensively –Construct intuitive tables •To avoid redundancy, keep your text poor in numbers –and rich in (simple) words

Short is beautiful (and wise)

A short text often contains much wisdom. –Sophocles To get the right word in the right place is a rare achievement. Anybody can have ideas - the difficulty is to express them without squandering a quire of paper on an idea that ought to be reduced to one glittering paragraph. –Mark Twain

… and Ezra Pound: • The first test of any writer is how many useless words he uses –i.e.: – how many words do not contribute to the point, – how many words detract from the most important point • Don’t be afraid of a short paper – remember Watson & Crick, Nature 1953, on DNA structure: a single page • Short sentences often reveal (and cure) problems in logic Tassios –English usage

Simple words

…simple language, short words and brief sentences. That is the way to write English… when you catch an adjective, kill it (not utterly, but kill most of them – then the rest will be valuable). –Mark Twain

•Avoid acronyms •Avoid jargon •Avoid many words when one will do

The right words

Science strives to make the new intelligible in terms of the familiar. – Friedrich Nietzsche

•Use scientific terminology accurately •Avoid colloquial expressions •Avoid cliches •Distinguish yourself and attract readers by expressing your own thoughts in a direct, lively manner

Repetition is not redundant

•Saying the same thing more than once is redundant –to be avoided •Using the same word whenever it is the most accurate one increases clarity –to be adopted Tassios –English usage

Check for logic

First –and last!

Critical (re)reading

•Sometimes what seems very important to the author is less important to the reader – •Ask colleagues (non-specialists if possible) to review your final manuscript (and of course acknowledge their effort) •Careless mistakes make a bad impression; speed up the publishing process by correcting before submission.

Conclusions

•Treat your results with the respect they deserve •Write about your work simply, clearly, without clichés •However, do not take your work so seriously that you fail to read your manuscript with the eye and attitude of your intended audience: –the editors and reviewers who will decide whether it is published, –your peers who will eventually read (and cite) it Tassios –English usage

Acknowledgment

Many thanks to Judith Crane, whose contribution to these notes was invaluable.

Appendix to Presentation 4 English usage: simple language for complicated matters

Documents: · J D Watson & F H C Crick (1953): Molecular Structure of Nucleic Acids: A Structure for Desoxyribose Nucleic Acid · M Ptashne (2007): Words · M Ptashne (2007): On Speaking, Writing and Inspiration · M Ptashne (2007): On Learning to Write original papers

there was no ‘activation’ that he for these various cases I don’t My Word could see. Only later did I realize know it. ‘Transactivate’, a term the problem: the word ‘activation’ sometimes used, adds nothing suggests just what does not but fog. Words happen when Gal4 (like many other so-called transcriptional Cooperativity Mark Ptashne activators in bacteria and in Here a problem can arise from higher organisms) works. exposure to a classical scientific “Every word is a prejudice” The word ‘activation’ suggests education. A famous example of a conversion from an inert cooperativity involves the binding Turn on a TV news channel to an active state – like what of O2 to haemoglobin. As we in the US and you are likely happens when you turn a key learned once, haemoglobin binds to see some well-spoken, and start a Ferrari engine. four O2 molecules, and binding attractive personality referring Some activators — cyclins, for of each O2 increases the affinity to opponents of the Iraq war example — cause their targets of the haemoglobin for the next as members of the ‘hard left’. A (certain kinases in this case) to O2. Binding of O2 changes the large fraction of the American undergo a transition from an shape of the protein (or if one public opposes the war, of inert to an enymatically active prefers, traps the haemoglobin in course, but we are inured to state. But Gal4 doesn’t literally a conformation different from the this prejudicial use of language. ‘activate’ anything, neither O2-free form) and this alternative At least one network speaks the enzyme (RNA polymerase) form of haemoglobin binds quite purposefully, but one nor the gene (whatever that subsequent O2 molecules with wonders about the others — are might mean). Rather, it simply increased affinity. We say that the the announcers themselves, recruits (a word discussed more binding of one O2 ‘helps’ another along with the rest of us, being fully below) the transcriptional to bind to haemoglobin. We manipulated by their own words? machinery (which includes RNA are so used to this description Scientists think of themselves polymerase) to the gene. This that sometimes the word as being in control of the apposition causes the gene to be ‘cooperativity’ is automatically words they use rather than the transcribed at a higher rate than associated with a required other way around. Nietzsche’s that observed in the absence of conformational change. comment above might give us the activator. The mechanism is But there is another example pause. I have discussed in a so simple as to be elusive. of cooperativity, widely used in previous communiqué unwitting This problem with the word Nature, that requires no changes problems engendered by activation applies to a wide array in the shapes of the components. too- facile use of the scientific of biological control processes The formal description of the term ‘epigenetic’ [1]. Here my because the relevant mechanism reaction is the same as that concern is with four rather in these disparate cases is of the haemoglobin case: one ordinary words — ‘activation’, essentially the same. In all of ligand ‘helps’ another bind to a ‘cooperativity’, ‘recruitment’ and these cases the enzyme (for common target. But here we are ‘regulation’. We have gotten so example, RNA polymerase) has dealing with macromolecules, used to using these words in one multiple possible targets (for for example, two proteins (the context — thereby engraining example, genes), and ‘activation’ ligands) binding to their affined our ‘prejudices’ — that we can means apposing the enzyme with sites on DNA. In this case, the be fooled when they are used one or another specific substrate. helping effect requires simply in another. In the context I will Just as a transcriptional activator that the two proteins touch emphasize, these words are apposes the polymerase with one another when bound to interrelated, and understanding a specific gene, so does, for DNA. More than once I have one requires understanding all. example, a subunit of an ‘E3 failed to explain this idea, only ligase’ appose a specific protein later to realize that the listener Activation with the ubiquitylating machinery associated ‘cooperativity’ with The yeast protein Gal4 is called (that which attaches ubiquitin to ‘conformational changes’ . The a ‘transcriptional activator’. It specific proteins and so marks idea here sometimes seems binds to specific sites on DNA them for proteolysis). There are too simple to grab the serious and causes the adjacent genes many such subunits (for example, person’s attention. to be transcribed, at a high level, F-box proteins) just as there are The source of the cooperativity into mRNA. We say that Gal4 many transcriptional activators, is easier to see in the case ‘activates’ transcription of these and their modes of action are involving three macromolecules genes. Some years ago, just strictly analogous. To say that an than in the case of haemoglobin when we had figured out how E3 ligase (in itself a misleading and O2. For example, for the Gal4 works, I showed a Scientific name) ‘activates’ ubiquytylation case of two proteins binding American article describing this is as misleading as to say that cooperatively to DNA, to a mechanism to a physicist friend Gal4 ‘activates’ transcription. first approximation one simply who, after a puzzlement, said that If there is a short alternative adds together three binding he simply couldn’t understand it: word to use in place of ‘activate’ energies: two protein-DNA and Current Biology Vol 17 No 14 R

required for recruitment are just like those required for cooperative binding –simple binding surfaces. Recruitment Enzyme Recruiter typically differs from cooperative binding only trivially: in the case of transcription, for example, the activator might be constitutively bound to DNA (as is the case for Gal4) and, only upon the appropriate signal would its ‘activating region’ (that which Substrate binds the transcriptional machinery) become exposed, and recruitment ensue. Recruitment is not synonymous with ‘increasing local concentration’: as the figure suggests there might be ways to bring the enzyme near its target (hence ‘increasing the local concentration’) in an inappropriate orientation. And in the eukaryotic world Current Biology especially, the recruiting reaction can be more elaborate than is Figure 1. Activation. A recruiter and an enzyme are shown binding cooperatively to a suggested by the figure. For common target, with three crucial pairs of interacting sites highlighted. example, some 50 proteins, As discussed in the text, the recruiter might be an E3 ligase, a transcriptional activator, found in an array of complexes, a splicing regulator, and so on. The recruiter might, as explained, be pre-bound, but the must be recruited to a yeast recruiting reaction remains essentially the same. Absent a recruiter, unless prevented gene to elicit transcription. from doing so, the enzyme will naturally work on its substrates at a lower level as dic- Some of these proteins are tated by its unaided affinity for its target.. recruited directly — are touched by the activator — and others one protein- protein interaction. specific substrate. Note the then bind cooperatively with Even a weak interaction between three pairs of interacting those directly recruited. Some the proteins — say a kcal or surfaces: substrate– recruiter; proteins (including even an two — can increase binding by a recruiter– enzyme; and inert RNA polymerase) might factor of 10–100, a large effect in enzyme– substrate. Were the be pre-bound to a gene, but in physiological terms. enzyme Eschericia coli RNA each of these cases the job of polymerase, the recruiter the activator is to recruit, by a Recruitment could be the bacterial protein simple binding reaction, some Alex Gann and I used lambda repressor (working as necessary component to trigger ‘recruitment’ to describe how an activator); and the substrate transcription initiation, to foster certain transcriptional activators, would be DNA bearing two transcriptional elongation, and such as Gal4, work [1]. As I have sites: an operator for binding so on. mentioned, and will explore repressor, and a promoter further in the next section, the for the polymerase. The Regulation word (and the mechanism) figure, with different names Our discussion of recruitment applies to an array of important inserted, equally well describes suggests one appropriate way to regulatory processes [2]. recruitment of a specific use this word. As we have noted, ‘Recruitment’ has been criticized protein to the ubiquitylating the cell contains an array of and sometimes misconstrued: machinery; recruitment of the active enzymes — polymerases, it has a ‘militaristic’ aspect, we dosage compensation complex ubiquitylators, proteases, RNA have heard; or it implies some to X chromosomes in one sex splicers, histone modifiers – each ‘molecular beacon’ effect; or of Caenorhabditis elegans; of which has multiple possible it specifies some particular recruitment of the RNA splicing targets. Each of these enzymes interaction (for example machinery to specific RNA sites, is regulated, then, by directing it recruiter–polymerase) that might and so on. to one or another of its possible not actually obtain for certain All recruiters of a given class substrates — a specific gene cases. We think the word apt, (transcriptional activators, for or protein, for example. I have however, and we see no obvious example) use one surface to emphasized the use of protein alternative. bind to the enzymatic machinery recruiters (such as transcriptional Figure 1 shows a protein and another to bind to a specific activators and E3 ligases) to recruiting an enzyme to a substrate. The kinds of surfaces effect this regulation, but other Magazine R kinds of molecules can also work as recruiters. Double-stranded RNA, for example, directs the RNA interference machinery to specific sequences in mRNA and in DNA, and thereby regulates a form of gene silencing. Signals typically are conveyed to recruiters and, as implied by the discussion thus far, not to the enzymatic machineries themselves. For example, the sugar galactose (which we’ll call a signal) causes the inhibitor bound to Gal 4 (called Gal80) to dissociate from Gal4, thereby triggering recruitment of the transcription machinery. This rule (that signals go to recruiters) is not ironclad: during formation of germ cells in Drosophila, for example, the transcriptional machinery is turned off entirely, and it would seem proper to call this a form of regulation. Despite our lack of a precise definition, there are uses of the word ‘regulation’ that are inappropriate. For example, let us say that RNA polymerase, once recruited along with whatever else is required for transcription, undergoes some conformational change as it begins to work — surely it does. But to call such a conformational change ‘regulatory’, without any evidence that such a step is subject to modulation by changing signals in the cell, is to embark on a trail of endless regress in which every event on a biochemical pathway can be called ‘regulatory’. This discussion has centered on ‘activation’, the imposition of specificity of by recruitment. ‘Repression’, as we shall see in our next encounter, is often, especially in eukaryotes, another manifestation of recruitment.

References Ptashne, M., and Gann A. (1997). Transcriptional activation by recruitment. Nature 386, 569–577. Ptashne, M., and Gann A. (2002). Genes and Signals, Cold Spring Harbor Press, Cold Spring Harbor, New York.

Sloan Kettering Institute, Memorial Sloan Kettering Cancer Center, 1275 York Avenue, Box 595, New York, New York 10021, USA. E-mail: [email protected] Magazine R

by the master to the neophyte in a good guide. Arnold Steinhardt, My Word Stephen Potter’s Lifesmanship. the violinist, recounts that after the Having spied the master during inaugural performance by his string the first half of a concert, the quartet (the Guarnari) they were On Speaking, neophyte approached him later at visited backstage by Rudolf Serkin. the pub, and challenged as follows: The pianist said “The last time I Writing and “The Debussy (in the second half) heard Mozart played like that was Inspiration was not good, don’t you think?” in Vienna — and that is why I left”. To which the master replied with And perhaps the hardest lesson astonishment: “You mean you to accept is that, unless it’s a thesis Mark Ptashne stayed for the Debussy?” exam or something of the sort, I wish people wouldn’t worry your audience really doesn’t care A visitor, giving a seminar at Cal about being too transparent. I once how hard you worked to get to Tech, found to his amazement asked Francis Crick why he spent your answer. Frank Stahl once said that when he had finished his the day with a pile of Scientific to me that most experiments are introduction Max Delbrück raised American magazines and he said just forays to teach you how to do his hand and said “Stop!, say “When you are learning something the right one. One good experiment it again”. “Say what again?” new the hardest thing is to get the is worth ten messy ones. So if asked the speaker, and Max said basic idea.” The truth is, most of you have something to say, say it “Everything you just said”. So he the time I’d prefer to hear what the simply and directly — you’ll make repeated, word for word, his 15 speaker thinks than what he can a better impression than if you feel minute introduction at which point prove — the proof can come later. obligated to say everything and Max raised his hand, said “Stop. I’ve never attended a seminar in bits of nothing all at one time. That’s what I thought you said”, which there were too few slides, “Science strives to make the new and walked out. or the slides were too simple, or intelligible in terms of the familiar” What prompted Max — who the speaker failed to use enough —Nietszche. But there is no lingua imposed his formidable will and technical terms, or — amazingly franca, and everyday terms mean intellect on the nascent field of — spoke for too brief a time. one thing to some of us, and molecular biology — to walk out? Aaron Novick told me, long ago, something else to others. Writing My guess is that the lecturer didn’t that I had to go to meetings and is even harder than speaking say anything wrong so much as seminars — only by looking at the – unless you are a master, there he didn’t say anything at all — at speaker, he said, would I know who are no rhythms, inflections, and least not clearly. I go to seminars was believable. That was back in half sentences to steer the reader surreptitiously, if possible, and the days when there was only a along. There is the constant tension only if there is an inconspicuous handful of potentially interesting between being communicative escape route. The danger is people. and being strictly correct — we that the speaker will begin by Let me put the matter this way. sometimes have to ‘lie the truth’. bandying about some key terms, I recently heard a seminar I loved. This does not mean that the best showing a few bewildering slides, The young woman sailed along of us speak or write the same and referring to all this stuff we briskly, every sentence having a way. I think of two inspirations of supposedly already know. I find point without being pedantic — opposite styles: Francois Jacob myself struck by a sentence, there was a salient quality of and Jacques Monod on one hand, wondering what exactly was mind. I was reminded: a seminar and Al Hershey on the other. meant; what constructs is this is a performance that has to be Jacob and Monod, Cartesians person carrying around in his rehearsed (even if silently, in bits) at heart, seemed to invent the head? And by the time I emerge over an extended period. It is not world before they stooped to from my musings the seminar just a matter of choosing which discover it; whereas Hershey took is half over, all is lost, and I words to use — equally important us through the nitty gritty — such sheepishly extricate myself. This is the choice of which words not as the drama of DNA folding and is risky — I have to rely on friends to use. In an otherwise good talk, unfolding — as though it were who can sit in the fog to extract speaking a wrong word or phrase happening before our very eyes, whatever of importance might be can be a disaster — you or the and thereby revealed a world. I there. audience will be diverted into would read both with exhilaration, Concerts too — I try to sit at explanations (or puzzlements) you thinking that what I wanted was to the end of rows to facilitate early desperately want to avoid. Clear do some experiment that would getaways when required; I can thinking does not guarantee a good enable me to find a voice so that I always puzzle over the review talk: Matt Meselson told me about could write — speak — like that. the next day. In ‘Old Vienna’, so a well-known scientist who, giving the story goes, there was a cost a seminar, gave the impression that Sloan Kettering Institute, Memorial Sloan for such anti- social behavior: a recording of a perfectly coherent Kettering Cancer Center, 1275 York Avenue, Box 595, New York, New York entrance to concerts was free talk was being played in his head, 10021, USA. but there was a charge if you left and he was commenting on it as it E-mail: [email protected] early. I am reminded, in a further went by. Unfortunately the way ‘it digression, of a lesson taught used to be done’ is not necessarily Magazine R1

produced! And what a valuable then most lines thereafter crossed My Word lesson it was. out. Madame Auclair, the French There are some rules that help, violinist, had a gentler approach. I suppose: short sentences, the Out of central casting, as they say: On Learning to active voice, as few technical and dark glasses, cigarette dangling, compound words as possible, hoarse , accented voice. A friend Write and so on. I used to write by went for a special violin lesson, hand, read (out loud) into a and asked whether he might tape Mark Ptashne tape recorder, re-read the typed record this important event in his outcome, throw away, read a page life. “Of course, my boy.” He played I learn from Bill Bryson in his fine of Nietzsche, and start again. But a bit and she said: “Very nice. There “A Short History of Everything” in my experience these rules and are some good things about your that obscure scientific writing methods are only the starting point, playing, very good. Now turn off the has a well-established history. and some rather more ‘interactive’ tape recorder”. Newton wrote impenetrably to instruction is required. Watson When I am struggling over keep tourists out; the geologist applied the following method (at yet another of my obscurely Hutton, with profound things least to me): my finely honed draft written drafts I sometimes recall: to say, wrote obscurely, and to was sailed back across the table amateurs play music ‘in general’; his detriment, because he was accompanied by an eyebrow- professionals play each note. And incapable of writing a coherent push-up-grimace and the word: so I present to a tough-minded English sentence. Is there a third, “Unreadable”. Reminds me of my friend one paragraph — just one perhaps more modern, category all-time most memorable violin — and when that is reported to be of obscurantism? Let’s face it: day lesson. I walked into the home of transparent I go on to the rest. But by day molecular biology can be the Russian virtuoso with whom even if I have followed the rules I tedious stuff indeed. Surely there I was studying and he said, mentioned above, and even if that must be some better world — one deadpan: “I see you are smiling. first paragraph seemed fine at the in which we solve meta-problems Why are you smiling? If I played like time, now, in view of what else I rather than quotidian ones. The you I wouldn’t smile.” have written, that first paragraph public (including us) wants what we We call this the boulder-in-the- might have to go, or be seriously might call a ‘leapfrogging’: there road teaching method, and it is not recast. Each paragraph is an must be a way, or ways, to describe so uncommon in music, especially experiment — you might not know and deduce in general rather than among the great Russians. A friend for some time whether it is any in particular. In that breathless of mine tells me about her friend good. atmosphere, obscurity can be who went to study with Heifetz There is a theme here, beautifully useful to the writer, ignored or — yes, Jascha Heifetz. The first expressed by a friend who was encouraged by editors, embraced week he told Heifetz he would play going through the agonies of the by administrators and deep the Sibelius Concerto, and JH said “ just the first paragraph” method pockets. “We’ll see about that”. The student in attempting to re-write a book. I Lets put these dark thought got through the first page before hadn’t heard from him in a while aside. Most of us, I’ll assume, do being sent home “to practice”. and began to worry — had I been want to write clearly, but how does Upon repeated attempts he never too tough? — and he wrote: “The one learn to do it? Our brains (some got past the first page until, the last only reason I hadn’t sent it (the new anyway) work in fits and starts — day, JH let him play the whole thing paragraph) already is that I didn’t this reminds me of that, that and then said: “It’s as I thought: you want to disappoint you. But I realize reminds me of this, do you know can’t play the Sibelius Concerto. that the only way you can help me the joke about Sam, and so on. Next.” Before being too harsh is if I continue to disappoint you. But standard scientific writing on Jascha, recall the story about So here it is...” All my teachers, won’t allow that to be transferred Max Delbrück — the very Max I whatever their methods, were trying to the page. Some people think mentioned in my last communiqué to help me, and I love them for it. the problem must be tackled on [1]. It is said that Max returned a Heifetz I wouldn’t be so sure about. a grand scale: the student is told manuscript, torn to pieces, along Rules are one thing, but in the end to present and defend, in writing, with a note that said: “Please communication is all: at the end an experimental plan to solve an switch fields”. of a pleasant interview with a fine outstanding problem outside his Al Hershey didn’t bother to tear scientist of foreign extraction she main area of interest. Wow — a up my manuscript. I wrote a 20 shook my hand and said: “Its been job for JBS Haldane, but not for page paper for him and got it back a pleasure talking to me”. most of us. As a graduate student, with most lines crossed out and I assisted in a course taught by the occasional phrase circled and References. Jim Watson at Harvard. One of marked “Good”. So I rewrote and 1. Ptashne, M. (2007), On speaking, writing Watson’s requirements was that rewrote and it came back with not and inspiration. Curr. Biol. May 15th issue. each student write a three page a mark on the first page! Not a Sloan Kettering Institute, Memorial Sloan paper on something, anything, mark on the second! Then the third Kettering Cancer Center, 1275 York related to the course. Oh what page: a line through the middle, a Avenue, Box 595, New York, New York masterpieces of indirection were penciled-in “START HERE”, and 10021, USA. E-mail: m-ptashne@mskcc. Current Biology Vol 17 No 11 R2

org CONCLUSION

Scientific writing: how to go about it

Suggested resources

How to publish in Biomedicine: 500 tips for success

Jane Fraser

Getting Research Published: an A to Z of publication strategy

Elizabeth Wagner

Winning the Publications Game: how to write a scientific paper without neglecting your patients

Tim Albert

Advice

Ancient

Write quickly and you will never write well; write well, and you will soon write quickly.

Marcus Fabius Quintilianus, c. 65 AD

More recent

Directions: write, read, rewrite.

Repeat steps two and three as needed.

Susan Sontag, 17.04.09, New York Times