Banner ad for the tech recruiting company Triplebyte: 'Triplebyte is building a background-blind screening process for hiring software engineers'

What Is The Morning Writing Effect?

Many writers anecdotally report they write best first thing early in the morning, apparently even if they are not morning people. Do they, and why? (psychology, survey)
created: 11 May 2011; modified: 14 Nov 2018; status: in progress; confidence: possible; importance: 4

Ericsson 1993 notes that many major writers or researchers prioritized writing by making it the first activity of their day, often getting up early in the morning. This is based largely on writers anecdotally reporting they write best first thing early in the morning, apparently even if they are not morning people, although there is some additional survey/software-logging evidence of morning writing being effective. I compile all the anecdotes of writers discussing their writing times I have come across thus far. Do they, and why?

Interviews with writers often touch on their writing process; the hope of the reader is, deep down, to learn how they do the things they do and perhaps the reader can do the same thing. For the most part, the lesson I’ve taken away from such profiles is that every writer is different and there do not seem to be many generalizable practices; for every writer that thrives on writing in longhand with goose quills on parchment, another is unable to think outside a computer text editor, or needs to inhale rotting bananas, or sharpen pencils, or write in a cork-lined room, or insist on a loud phonograph/party for inspiration.

But in The Role of Deliberate Practice, Ericsson 1993 (among others), Ericsson draws on some anecdotes and particular long-running & somewhat-standardized Paris Review interviews of famous writers to make some interesting points about the relative brevity of most writing sessions (perhaps not too surprising as the physical typing/writing is not the bottleneck) but also the timing of it typically in the morning:

The best data on sustained intellectual activity comes from financially independent authors. While completing a novel famous authors tend to write only for 4 hr during the morning, leaving the rest of the day for rest and recuperation ([Cowley, M. (Ed.). (1959). Writers at work: The Paris review interviews.]; [Plimpton, G. (Ed.). (1977). Writers at work: The Paris review. Interviews, second series.]). Hence successful authors, who can control their work habits and are motivated to optimize their productivity, limit their most important intellectual activity to a fixed daily amount when working on projects requiring long periods of time to complete…Biographies report that famous scientists such as Charles Darwin, (Erasmus Darwin, 1888), Pavlov (Babkin, 1949), Hans Selye (Selye, 1964), and B.F. Skinner (Skinner, 1983) adhered to a rigid daily schedule where the first major activity of each morning involved writing for a couple of hours. In a large questionnaire study of science and engineering faculty, Kellogg (1986) found that writing on articles occurred most frequently before lunch and that limiting writing sessions to a duration of 1-2 hr was related to higher reported productivity…In this regard, it is particularly interesting to examine the way in which famous authors allocate their time. These authors often retreat when they are ready to write a book and make writing their sole purpose. Almost without exception, they tend to schedule 3-4 hr of writing every morning and to spend the rest of the day on walking, correspondence, napping, and other less demanding activities (Cowley, 1959; Plimpton, 1977).

This was interesting to me because I generally do not write in the morning, so knowing that morning is better would be valuable to me, and because it confused me why it would be true. If you are a morning person, you should write in the morning, and vice-versa if you are an evening person. Why would you write when you are miserable? I was especially not a morning person when I was a teenager, and it certainly showed in my first period class at 8:20AM after waking up at 6AM; certainly I never noticed any hidden gift for writing novels manifesting, even when it was a literature class. (Although I did notice a hidden gift for completely forgetting anything said in the first period.)

Some of the post hoc explanations for why morning might be better make no sense. It is true people are less likely to interrupt you early in the morning; but they are less likely to interrupt you at midnight. It is true people can find time for writing by getting up before their job; but they can sacrifice the same amount of sleep to write by staying up later at night. It is true that the morning might not be a circadian nadir; but that’s not helpful to anyone who is an owl, who by definition is sluggish in the morning, and where does all this energy come from for walking or exercising or partying or researching in the evening, when not writing, if the writer is hopelessly fashed after the vicissitudes of the day? (If the secret of morning writing is merely the nigh-tautological if you’re a morning person who writes best in the morning, you should write in the morning, and if you’re an evening person, you should write in the evening, then it’s surely of no value - is there anyone who doesn’t already know whether they are more of a morning or evening person?)

Further, morning writing runs counter to the usual intellectual stereotype of great writers as rising late and being chronometric night owls. The owl chronotype is usually linked with creativity & intelligence while morning larks are considered less creative (but more industrious & favored in many contexts like schools, to the detriment of owls & especially teenagers), so one would expect the opposite: writers to report writing mostly in the evening. (And amusingly, the Wikipedia article on owls includes a list of dozens of writers & other creative types while the lark article is devoid.) What could be more writerly or bohemian than spending the day researching or enjoying nature or drinking to jazz at the club and then returning to one’s attic in the witching hours to author deathless verse? Nevertheless, great authors routinely rehearse the advantages of living like a farmer and rising with the sun to milk the Muses. (Cal Newport’s Deep Work book points this way, and there is even a writer self-help fad, The Miracle Morning, whose central gimmick is getting up early.)

Puzzled, I began noticing in author interviews or writings that when writing times were mentioned, it was indeed more, often than not, partially or entirely in the morning (sometimes disgustingly early like pre-dawn), rather than usually in the evening as expected, and it became startling when I ran into an exception like Ian Fleming or Winston Churchill, who wrote at all in the evening, or Brandon Sanderson or Robert Frost, who work entirely into the wee hours. I am, again, not a morning person but I forced myself up early a few days and skipped my usual email & news-reading routine to focus on writing, and darn if it didn’t seem to work and the writing was worth the price in afternoon circadian slumps. Still couldn’t make myself do it regularly, though.

Causes

If the morning writing effect is real (in the sense that successful writers do disproportionately write in the morning - which is still in doubt given the existing systematic survey evidence is limited to ordinary writers while the elite writers are represented so far purely by haphazardly-selected anecdotes), what is causing it?

  • One possibility is that there is some sort of ecological fallacy going on: it is possible that, creativity really is higher in owls at night, owls do not improve by writing in the morning, but the best authors are still larks (rather than owls as one would assume from the population-level correlation) and do benefit from writing in the morning (or at least aren’t hurt); and this is because larks have other advantages in becoming the best authors, perhaps related to sheer writing volume & consistent output. No matter how creative it is, an unwritten book is no good. Larks then could write fine at any time and would be overrepresented among the best authors either way, because writing time is confounded with other Conscientiousness-related attributes.

    Looking through the anecdotes so far, while it’s true that devotees of The Miracle Morning and others frequently claim to not be larks and struggle to reap the benefits of morning writing, the elite writers who happen to write in the morning (currently) do not mention major struggles with getting up early or focusing in the morning, implying that they may well all be larks in the first place!

    (Simonton’s equal-odds rule suggests that volume of writing output is much more important than it is usually given credit for being, and that writing or research is too random a process to permit siting down for several years and decide to bang out a beloved masterpiece: one can only try as many things as possible and be surprised when one turns out well, or happens to become a hit.)

    • A related but somewhat simpler possibility is that people are bad at scheduling and while late-night writing is no different than morning writing and just as effective in theory, people tend to choose to fill up their schedules, and repeatedly accidentally find themselves with too little time to write at night; every hour of evening writing that people do get done is just as effective as morning hours, but there are fewer such hours. Doing it in the morning is then simply a little trick to make sure that other obligations literally cannot come first.
  • Another possibility is that the day really does use up some sort of willpower or creativity: all the little things one does before the writing late in evening fill up one’s mind. There is nothing special about morning hours, they merely happen to be the conscious hours closest in time to sleep pushing the big reset button on the brain. If someone slept during the day & woke up at midnight, that person would then be best off writing at midnight, right after waking up, rather than 8 hours later in the morning, equivalent to their afternoon. (Tononi’s SHY theory of sleep would be a low-level neurobiological explanation along these lines.) These sorts of theories have a problem with the existence of authors who prefer to work at the end of their (subjective) day and are energized at night - why are they not just immune to what should be generic effects of biology/psychology, but positively energized?

    • A version of this thing building up/wearing out over the day is that it is related to ego depletion or decision fatigue1 or opportunity cost, where the increasing number of accomplished activities it becomes an excuse to write less - I had a busy day, I can take it easy tonight. - or one has difficulty truly focusing because there are so many other things which one could do (Kurzban’s opportunity cost model).
  • Yet another version might be that sleep itself is the key: sleep, aside from any resetting, is also responsible for memory formation and appears involved in unconscious processes of creativity.

    Sleep is a long time period in between phases of working, allowing for the incubation effect to operate, and the incubation effect may be particularly benefited by sleep. So, one wakes up primed to work on the next piece of writing (that one has likely been mulling a long time), and by instead puttering around making tea or breakfast, one dissipates the potential. In this model, instead of one’s writing potential gradually deteriorating over the course of the day as the mind fills up/willpower is used up, it falls sharply and then hits a baseline and perhaps follows the usual circadian rhythms thereafter with a nadir at siesta time etc.

    • Or, perhaps there is something special about the liminal half-sleep state, which makes fantasizing or imaging easier. One parallel we might draw is with the ancient connection between fiction writers and alcohol: writers are notorious for drinking, often to excess. Is there something about the depressant or loosening of inhibitions of alcohol which assists writing, which might also be reproduced in the morning? On the other hand, nonfiction writers like journalists or philosophers or scientists tend to be associated with stimulants, particularly nicotine, caffeine, and amphetamines (not to mention modafinil)2; while those, particularly amphetamines, are less associated with fiction writers.3 (This makes me wonder if there is a connection to another anomalous anecdotal phenomenon, the so-called alcohol afterglow effect, and if my poor LSD microdosing results reflect my own nonfiction tendencies.) This half-asleep explanation neatly explains why evening doesn’t work. It wouldn’t apply to falling asleep as there is an asymmetry: a half-asleep person in the morning who is writing is getting gradually more alert and spending the rest of the day awake, and can build on whatever mental seeds were planted; while a half-asleep person in the evening would frustrate sleep by trying to write, can’t write for long before falling asleep, and when they do fall asleep, would forget the preceding ~10 minutes.
    • Or, perhaps it is a lack of sleep: sleep deprivation can cause odd mental states including mania and loss of inhibitions, and there is a peculiar but seemingly very real phenomenon where acute sleep deprivation in people with major depressive disorder substantially temporarily relieves their symptoms (Meta-Analysis of the Antidepressant Effects of Acute Sleep Deprivation, Boland et al 2017). Many writers are melancholic, so early mornings, especially cross-chronotype, might be an inadvertent rediscovery of this to the extent that they shortchange sleep in order to get up. Perhaps most people are not in the throes of full MDD, but there might be a more mild effect. If the sleep-deprivation effect is the culprit, then writers who do this need to be cutting sleep considerably and the effects will be only temporary, since chronic sleep deprivation doesn’t help (and worsens cognition); this might also explain anecdotes where the person maintains that morning-writing works for them but they could only do it for a few days or once in a while - naturally, the more sleep deprivation the harder it is to get up, and as both the sleep deficit builds up & the anti-depressant effect disappears, they will find morning-writing increasingly useless and will stop. This might seem like an undesirable hypothesis but it still allows occasional benefits on carefully-chosen occasions, such as finishing or starting a novel.

Directions

There are a few things one could do to generate a little more data on this:

  1. systematically go through the Paris Review interviews and the similar GoodReads interviews to note down all cases where an author is asked about writing time, rather than a few examples; this avoids the risk that morning writing advocates have selectively chosen examples from the interviews. As the writers are not chosen for their writing habits and the interview question are fairly formulaic, presumably the interview series could be considered a quasi-random survey sample of successful authors.
  2. run a population-sample survey (I have done one USA survey myself but more extensive surveys & surveys elsewhere would be useful)

    • run surveys in more elite-writer samples
  3. run a (non-blinded) self-experiment: create a list of things to systematically work on; flip a coin to decide whether to get up early, record total words-written+time-spent etc. (I can’t decide if I would be biased towards wanting it to work or wanting it to fail: of course I want to be better at writing, but on the other hand - surely there’s some easier way!)

Research

  • Writing method and productivity of science and engineering faculty, Kellogg 1986: to go into more detail, it reports:

    The respondents tended to schedule their work between 8AM and 8PM, with the morning hours being the most common time of day (Table 3). Positive but nonsignificant correlations were obtained for these time intervals. Night owls were rare and not unique in their productivity. In terms of the duration of writing sessions, the data indicate a preference for one to three hours. Working for 1 to 2 hours was significantly correlated with productivity. But as will be explained later in describing the multiple regression analyses, this effect is best attributed to other factors correlated with the frequency of working for 1 to 2 hours. Highly regular work scheduling was not the rule; the most common response was only a 3 on the 7 point scale. Write in spurts and marathon writing just before a deadline were comments listed by respondents that match the pattern commonly observed in Boice and Johnson’s (op. cit.) survey. As in Boice and Johnson’s study, regular writing was positively correlated with productivity, but here the relationship was weak and nonsignificant.

    Survey Item Mean Mode Std. Dev. Productivity Correlation (r)
    Midnight-4AM (hour of day) 1.76 1.00 1.29 0.01
    4AM-8AM 1.87 1.00 1.49 0.04
    8AM-Noon 4.61 6.00 1.44 0.17
    Noon-4PM 4.34 4.00 1.33 0.15
    4PM-8PM 3.60 4.00 1.54 0.13
    8PM-Midnight 3.80 2.00 1.80 0.05
    0-1 hour (Duration) 3.50 2.0 1.58 0.09
    1-2 hours 4.46 6.0 1.40 0.22*
    2-3 hours 4.44 6.0 1.36 0.07
    3-4 hours 3.49 4.0 1.63 -0.04
    More than 4 hours 2.76 1.0 1.73 -0.12
    Every working day (regularity) 3.01 3.0 1.50 0.11

    Table 3: Analysis of Work Scheduling (n=121; The response scale ranged from Never (1) to Always (7). * = p<0.05)

    Another interesting aspect of Kellogg 1986 is that almost all variables correlate non-statistically-significantly with productivity (defined in Kellogg as the total number of books/papers/reports/grant-applications/grant-reports written in the previous 3 years), and most are of small magnitude. Measurement error & range restriction come to mind as biasing effects towards zero, but it’s still consistent with my own experience that it is difficult to find anything which strongly correlates with productivity, much less causes it.
  • The psychologist as wordsmith: a questionnaire study of the writing strategies of productive British psychologists, Hartley & Branthwaite 1989 conduct a similar survey as Kellogg but do not give any statistical details that I can find, saying merely

    In the present study most of our productive psychologists had no real preference for any time of day at which to work. The morning appeared to be slightly preferred to the afternoon and the afternoon slightly preferred to the evening. Regular working times were correlated with overall productivity, but productive book writers wrote sporadically (in term time). These findings were very similar to those of Kellogg (1986) who showed that the majority of his 121 engineers worked in the morning, and then the afternoon, but that a highly regular work schedule was not the rule.

  • Productivity in 2017: What we learned from analyzing 225 million hours of work time, RescueTime 2018; analytics over hundreds of thousands of users:

    Looking at the time spent in software development tools, our data paints a picture of a workday that doesn’t get going until the late morning and peaks between 2–6pm daily…While writers are more likely to be early birds…Writing apps were used more evenly throughout each day with the most productive writing time happening on Tuesdays at 10AM.

    Time spent in writing tools (light blue): RescueTime analysis of distributing of writing app use over time of day over the week: note intense band 10-11AM every day
    Time spent in writing tools (light blue): RescueTime analysis of distributing of writing app use over time of day over the week: note intense band 10-11AM every day
    Allowing for the different time buckets, the RescueTime results closely parallel Kellogg 1986’s survey responses. Aside from being an enormous data sample, RescueTime notes an interesting contrast: despite being apparently similar activities (both mostly involve slinging text), the temporal timing of software development & writing are strikingly different. Thinking back, I don’t recall early-morning programming being a trend among programmers (programmers are infamous for preferring to come in late and late-night programming sessions which may wrap around the clock, especially in college - though the original reason, that the computers are less busy at night, has long since expired). It’s fascinating that the stereotypes about writing vs programming line up so well with the RescueTime data.
  • 2018 Google Surveys, general USA population sample, asking self-identified writers/researchers/scientists their chronotype & ideal writing time

    I ran this survey in October 2018, using Google Surveys, asking a question akin to Kellogg 1986’s survey, like if you are a professional writer, blogger, researcher, or scientist do you find you write best at: [not a writer]/[Midnight-4AM]/[4-8AM]/etc? At $0.10 a response, if 5% of the population could be considered some sort of writer (which sounds reasonable to me) and we want another n=121 to equal Kellogg 1986’s sample size, the survey would only cost 0.1020121=$2420.10\cdot 20 \cdot 121 = \$242. A second question could be added to ask if the respondent considers themselves more of a morning or evening person, however, it dectuples the cost; as in my catnip surveys, it should be possible to combine the questions into a single question which can hopefully provide a useful datapoint. (GS tries for a representative population sample by techniques like reweighting; I don’t know if they take time-of-day into account and thus lark/owl type, but the surveys typically run over several days so hopefully they wind up being inherently balanced anyway.)

    If morning is the most common (replicating the Kellogg 1986 & RescueTime results), and if many evening-preferring respondents still answer that mornings work better for writing, that would be pretty good evidence for morning writing being a real phenomenon (although still leaving the causal status ambiguous and not answering the question of whether owls - like me - would benefit from switching to morning writing).

    On 27 October 2018, I launched an all-ages all-gender USA population survey on Google Surveys. Because of the need to run as a 1-question survey, I condensed the two questions into one and simplified it considerably into just morning/evening preference and morning/evening self-estimated writing performance, giving 5 possible responses (1+2x2=5). As most respondents will be useless - I guesstimate 5% would consider them professional writers of some sort, so for a few hundred responses, I need 20x as many; I settled on n=5000/$500 for the survey, which should deliver a precise enough result. The question looked as follows:

    1. If you are a writer/researcher/scientist, are you a morning/evening person & when do you write best? [answers displayed either ascending or descending at random]

      • I don’t write or blog
      • Morning person; best writing during morning
      • Morning person; best writing during evening
      • Evening person; best writing during morning
      • Evening person; best writing during evening

    The survey finished 29 October 2018 with the following results (percentage is population-weighted out of equivalent n=3,999; n is raw counts out of the n=5004 actually collected; CSV)

    1. 70% (3515)
    2. 9.9% (467)
    3. 4.0% (196)
    4. 4.0% (193)
    5. 12.1% (633)

    The percentage of people willing to claim to be writers was ~6x larger than I expected, which is troubling (do really that many people write large amounts?). Otherwise, the responses appear reasonably evenly balanced: 663 morning people vs 826 evening people. The percentage of overall counter-chronotype self-rated writing performance is 26%. On average, 55% of respondents thought evening-writing was best. The key question, of course, is whether morning-writing is more preferred for counter-chronotype writing: there is a slight preference here, but it is the opposite of predicted, with 29% of morning people believing they write best in evening versus 23% of evening people saying morning is best for them. (The difference is statistically-significant at p=0.008/P~1.4)

    This does not strongly endorse morning-writing, although it is surprising how many people think they write best counter-chronotype. Of course, the fact that fewer people believe they write better in the morning rather than evening doesn’t prove morning-writing isn’t a thing: one possibility is that people simply haven’t given it a fair try, or that it only works for professional writers at a high level, or that it is heterogeneous and there is a small fraction of people for whom morning-writing works really well (and so everyone should give it a try just in case). The overall even split of chronotype does give a baseline expectation for elite writers, though.

Anecdotes

Table

Author Date Type Time Hours
Kellogg survey 1986 Nonfiction Morning-afternoon 8AM-12PM, 12PM-4PM
RescueTime users 2018 Nonfiction Morning 10AM-11AM
Gwern Google Surveys 2018 Fiction+Nonfiction Evening ?
Kazuo Ishiguro 2014 Fiction Morning 9AM-10:30AM
Dan Brown 2017 Fiction Morning 4AM-12PM
Philip Pullman 2017 Fiction Morning-afternoon 10AM-1PM
Ian Fleming 1964 Fiction Morning-evening 10AM-12PM, 6-7PM
Joseph Campbell ? Nonfiction Morning-evening 9AM-6PM
Charles Dickens ? Fiction Morning-afternoon 9AM-2PM
Robert Frost ? Fiction Afternoon-evening 1PM?-3AM
Winston Churchill ? Nonfiction Morning-evening 11AM-1PM, 11PM-2AM
Frank Herbert 1969 Fiction Morning-evening 5AM-7AM, 5PM-1AM
Harry Harrison 1968 Fiction Afternoon 12:30PM-5PM
Toni Morrison 2015 Fiction Morning 6AM-10AM
Chimamanda N. Adichie 2013 Fiction Any Any
Michael Connelly 2017 Fiction Morning 4AM-7AM
Stephenie Meyer 2016 Fiction Evening 8PM-12PM
Stephen King 2014 Fiction Morning 8AM-12PM
Paulo Coelho 2014 Fiction evening ?PM-4AM
Brandon Sanderson 2012 Fiction Evening 12PM-4PM, 4PM-3AM
Margaret Atwood 2014 Fiction Morning-afternoon ?AM-?PM
Saint-Exupéry 1947 Fiction Evening 11PM-8AM
Neil Gaiman 2004 Fiction Evening ?PM-?AM
John Irving 1986 Fiction Morning-afternoon ?AM-?PM
Donald Hall 2018 Fiction Morning-afternoon 5AM-?PM
Hunter Thompson ? Nonfiction Evening 12AM?-6AM?
Michel Houellebecq 2018 Fiction Morning ?AM-?AM

Examples

Additional anecdotes of writers’ preferred time I’ve collected:

  • Kazuo Ishiguro (2014 interview): morning+evening

    I would, for a four-week period, ruthlessly clear my diary and go on what we somewhat mysteriously called a Crash. During the Crash, I would do nothing but write from 9AM to 10:30PM, Monday through Saturday. I’d get one hour off for lunch and two for dinner.

  • Dan Brown (profile): morning (>4AM-noon)

    Mr. Brown, 53, spent four years writing and researching the book. He is nothing if not disciplined. He rises at 4AM each day and prepares a smoothie comprising blueberries, spinach, banana, coconut water, chia seeds, hemp seeds and … what’s the other kind of seed? he asked. Flax seeds, and this sort of weird protein powder made out of peas. He also makes so-called bulletproof coffee, with butter and coconut oil, which he says changes the way your brain processes the caffeine so as to sharpen your mind. His computer is programmed to freeze for 60 seconds each hour, during which time Mr. Brown performs push-ups, situps and anything else he needs to do. Though he stops writing at noon, it’s hard for him to get the stories out of his head. It’s madness, he said of his characters. They talk to you all day.

  • Philip Pullman (profile): late morning-early afternoon (10AM-1PM)

    Every day from roughly 10 until 1, Pullman sits at his desk in a monkish study at the top of the house and produces three pages, longhand. He has written three pages a day ever since he started writing. Habit, he is fond of saying, has written far more books than talent. The ritual is sacred. As is the space. Nobody’s photographed this, and nobody will ever photograph this, he told me, both fierce and faintly amused by the severity of his own rule. I’m superstitious about that, very superstitious about that.…For a man whose novels are restless, whose characters never stop traveling, Pullman leads a relatively static life. After the morning shift at his desk, he spends his afternoons either tending to the 800-odd trees he and Judith have planted in the fields behind their house or in his carpentry workshop, where he makes things like reading stands and chopsticks. Occasionally, he drives an elderly woman in the village to the library, and he goes to the cinema once a week with his publisher and close friend David Fickling and their wives. I have the company of the people I’m writing about, Pullman told me. Jude and I are quite happy here with our hermit-like existence.

  • Ian Fleming (Playboy Interviews II, December 1964 interview; pg56-57): morning+evening (10AM-noon, 6-7PM)

    Playboy: Do you spend most of your time there at the typewriter?

    Ian Fleming: “By no means. I get up with the birds, which is about half-past 7, because they wake one up, and then I go and bathe in the ocean before breakfast. We don’t have to wear a swimsuit there, because it’s so private; my wife and I bathe and swim a hundred yards or so and come back and have a marvelous proper breakfast with some splendid scrambled eggs made by my housekeeper, who’s particularly good at them, and then I sit out in the garden to get a sunburn until about 10. Only then do I set to work. I sit in my bedroom and type about fifteen hundred words straightaway, without looking back on what I wrote the day before. I have more or less thought out what I’m going to write, and, in any case, even if I make a lot of mistakes, I think, well, hell, when the book’s finished I can change it all. I think the main thing is to write fast and cursively in order to get narrative speed.

    Then, about quarter-past 12, I chuck that and go down, with a snorkel and a spear, around the reefs looking for lobsters or whatever there may be, sometimes find them, sometimes don’t, and then I come back, I have a couple of pink gins, and we have a very good lunch, ordinary Jamaican food, and I have a siesta from about half-past 2 until 4. Then I sit again in the garden for about an hour or so, have another swim, and then I spend from 6 to 7 - the dusk comes very suddenly in Jamaica; at 6 o’clock it suddenly gets very dark - doing another five hundred words. I then number the pages, of which by that time there are about seven, put them away 1n a folder, and have a couple of powerful drinks, then dinner, occasionally a game of Scrabble with my wife - at which she thinks she is very much better than I am, but I know I’m the best - and straight off to bed and into a dead sleep."

    Playboy: And you return to England in March with a completed manuscript?

    Ian Fleming: Except for minor revisions, yes.

  • Joseph Campbell (quoted from The Hero’s Journey: Joseph Campbell on His Life and Work): researched/studied in multiple blocks morning-evening

    So during the years of the Depression I had arranged a schedule for myself. When you don’t have a job or anyone to tell you what to do, you’ve got to fix one for yourself. I divided the day into 4 four-hour periods, of which I would be reading in three of the four-hour periods, and free one of them. By getting up a 8 o’clock in the morning, by 9 I could sit down to read. That meant that I used the first hour to prepare my own breakfast and take care of the house and put things together in whatever shack I happened to be living in at the time. Then three hours of that first four-hour period went to reading. Then came an hour break for lunch and another three-hour unit. And then comes the optional next section. It should normally be three hours of reading and then an hour out for dinner and then three hours free and an hour getting to bed so I’m in bed by 12. On the other hand, if I were invited out for cocktails or something like that, then I would put the work hour in the evening and the play hour in the afternoon. It worked very well. I would get nine hours of sheer reading done in a day. And this went on for five years straight. You get a lot done in that time.

  • Charles Dickens (based on Charles Dickens: A Life, Tomalin 2011, cited by McCrum): morning-afternoon (9AM-2PM)

    But then, as you go deeper into Tomalin, you discover that Dickens, in his prime, used to compress his literary energies into five hours, roughly 9am to 2pm, after which he would walk incessantly, and put his mind into neutral. He might return to what he’d written in the morning later in the evening, but those five hours held the key to his output.

  • Robert Frost (McCrum): afternoon-late evening (?-3AM)

    Robert Frost, whose remote Vermont cabin I visited recently in company with his biographer Jay Parini, never started work till the afternoon, and often stayed up till two or three in the morning, not rising until midday, or even later.

  • Winston Churchill (as cited to The Last Lion, Manchester): morning+late evening (~11AM-1:15PM, 11PM-2AM)

    Once all the newspapers have been perused, it’s time to answer the enormous amount of mail Churchill receives each day. A secretary stands by as Churchill dictates (his preferred method of writing) correspondence to private citizens and government officials. Once the mail is finished, it’s time to dictate memoranda and greet any visitors who have stopped by Chartwell. He will receive anyone except the King in his bedchamber, and visitors are often tickled by the image which greets them; Vice Admiral Sir Douglas Brownrigg said he presented a most extraordinary spectacle, perched up in a huge bed, with the whole of the counterpane littered with dispatch boxes, red and all colours, and a stenographer sitting at the foot - Mr. Churchill himself with an enormous Corona in his mouth. Churchill’s next task is to look through galley proofs of the latest book he’s working on, and ask his chief researcher to check and verify certain details. At this point, he often begins to work on his speeches. He paces the room, issuing phrase after phrase at a speed his secretaries have trouble keeping up with. Churchill, one of them recalls, would be dashing around in shorts and undershirt and a bright red cummerbund while I trotted behind him from room to room with a pad and pencil struggling to keep pace with the torrential flow of words. This flow of masterful oratory increases as the wordsmith warms up and finds his groove; By noon the cadences of his prose have begun to trot; by 1:00PM they are galloping. Lunch is at 1:15, so Churchill sets aside business and gets dressed to the nines (hence the aforementioned cummerbund)…Churchill believes his afternoon naps help him be much more productive. He has found that he can only produce good writing for a few hours at a stretch, before his brain gets tired and the quality diminishes. So by breaking up his schedule with a nap, he is able to have two creative working periods each day - one in the morning and one late at night - while also having time for socializing and duck feeding…The guests have gone home or retired to their bedrooms to stay over, Churchill begins his second working shift of the day. It’s 11:00 PM, and most of his fellow Englishmen are sleeping, but Churchill is rearing to go. He slips into something more comfortable and asks his aides to join him in the library:

    His appearance heralded by the harff, harff of his slippers, he enters the room in his scarlet, green, and gold dressing gown, the cords trailing behind him. Before greeting his researcher and the two secretaries on duty tonight, he must read the manuscript he dictated the previous evening and then revise the latest galleys, which arrived a few hours earlier from London. Since Churchill’s squiggled red changes exceed the copy set - the proofs look as though several spiders stained in crimson ink wandered across the pages - his printers’ bills are shocking. But the expense is offset by his extraordinary fluency. Before the night is out, he will have dictated between four thousand and five thousand words. On weekends he may exceed ten thousand words.

    Churchill’s night usually ends around 2 am, but when there is extra work to be done, he may not retire until 3 or 4.

  • Frank Herbert (3 February 1969 interview with Willis McNelly): morning+evening (5PM-1AM and 5AM-7AM)

    Willis McNelly: What is your writing schedule?

    Frank Herbert: Well, it varies…depends on what I’m doing…writing for the magazine…but as a general rule it goes like this: I’ll get home somewhere around 5 o’clock when Bev is here, when she’s not working as she has been the last couple of weeks. She’ll have dinner ready at that time or very close to that time. I’ll then take an hour’s nap and then work sometimes until 1 o’clock in the morning. Then I hit the sack and get up and sometimes if a story is strong in me I get up in the morning and write…get up at 5 o’clock in the morning or so and write for an hour or two sometimes before going down to San Francisco.

    WM: Yes.

    FH: And this is the thing I want to get out of because I can write 8 hours a day in 2 bursts and I don’t see any reason why I shouldn’t be doing what I want…writing what I want to write during those times. I don’t envision supporting myself entirely by science fiction writing in the sense of writing only science fiction, because I have other axes to grind, too.

  • Harry Harrison (McNelly interview with Frank Herbert): afternoon (12:30PM-5PM)

    Willis McNelly: It’s interesting…Harry Harrison describes the writing process with him rather well in a tape I made with him a few months ago. He is absolutely uninterruptible from, say, 12:30 in the afternoon ’til 5:00 at night, because the ideas as they form in his mind sort of becomes extensions of his [cough] excuse me, fingers in his typewriter and that they are up here and that …that any interruption, whether it be a telephone ringing or his wife knocking at the door or anything at all is liable to shatter that idea as it transforms itself into paper.

  • Toni Morrison (Goodreads interview): early morning (6AM-10AM?)

    Goodreads: What’s your average writing day like? When do you write?

    Toni Morrison: Very early in the morning, before the sun comes up. Because I’m very smart at that time of day. Now, at this time of day [4PM], it’s all drifting away. But tomorrow morning I will be sharp for about 4 hours, say from 6AM to 10AM. If I get up before the sun and greet it, that’s when I start.

  • Chimamanda Ngozi Adichie (Goodreads interview): no particular time

    I wrote the book in both Nigeria and the U.S. I don’t have a routine. I like silence and space whenever and wherever I can get it. When the writing is going well, I’m obsessive - I roll out of bed and go to work. I write and rewrite a lot and shut everything out. When it is not going well, I sink into a dark place and read books I love.

  • Michael Connelly (Goodreads interview): very early morning when possible (4AM-6AM? pre-dawn)

    Goodreads: What’s your writing process?

    Michael Connelly: Because of working on a TV show [Bosch], my writing process is to write whenever I get a chance. Also, my training in journalism has taught me to write - I don’t need to be coddled. I can write in my office, I can write on planes, I can write in cars. I was on a plane last night for five hours, squeezed in so tight, my elbows were pushing into my ribs, but I wrote the whole time and got a lot done. That’s my process: to try to write whenever I can. A perfect day would be to get up before the light gets up in the sky and start writing and get a lot done before the rest of the city wakes up. That’s what I try to do when I’m at home or even when I’m in a hotel on the road. Morning hours are really good for me, dark morning hours. So in that regard I kind of share something with Renée [The Late Show character] because I like to work till dawn.

  • Stephenie Meyer (Goodreads interview): evening (8PM-?)

    Goodreads: Do you have any writing rituals?

    Stephenie Meyer: None, really, besides time of day. I can never get truly immersed in writing during the daytime. I know it’s a product of being interrupted by work calls and emails, children’s and husband’s questions about where fill-in-the-blank is located, and the dog’s bladder needs. Subconsciously my brain believes that there is no point in trying to focus when my office door is just about to slam open in three…two…one…. So now, even when I’m in a quiet, private environment, I can’t make my brain accept that it is possible to write while the sun is out. When I’m in the middle of a story, I do my self-editing during the day. That part handles interruptions better.

    2013 interview clarifies that darkness means late night, not early morning pre-dawn:

    That vivid scene took place in the kitchen but these days, she writes in an office room in her house. Like most writers, Stephenie prefers to write in the evening. I write best at night, which is one of the reasons that I’m so slow now, she said. My kids’ school schedule has forced me to be a morning person, which I am not. I don’t get so much done during the day. There are too many interruptions. It’s better if I can start at 8PM and work through till I pass out. Laughing, she quipped, Maybe, I should move back into the kitchen.

    In an autobiographical piece, Meyers says that the idea for Twilight came in a dream & was initially written in the morning but the bulk was written at night:

    I woke up (on that June 2nd) from a very vivid dream…Though I had a million things to do (i.e. making breakfast for hungry children, dressing and changing the diapers of said children, finding the swimsuits that no one ever puts away in the right place, etc.), I stayed in bed, thinking about the dream. I was so intrigued by the nameless couple’s story that I hated the idea of forgetting it; it was the kind of dream that makes you want to call your friend and bore her with a detailed description. (Also, the vampire was just so darned good-looking, that I didn’t want to lose the mental image.) Unwillingly, I eventually got up and did the immediate necessities, and then put everything that I possibly could on the back burner and sat down at the computer to write-something I hadn’t done in so long that I wondered why I was bothering. But I didn’t want to lose the dream, so I typed out as much as I could remember, calling the characters he and she. From that point on, not one day passed that I did not write something. On bad days, I would only type out a page or two; on good days, I would finish a chapter and then some. I mostly wrote at night, after the kids were asleep so that I could concentrate for longer than five minutes without being interrupted.

  • Stephen King (Goodreads interview): morning (8AM-noon)

    Goodreads: What’s your average working day like? Do you have any unusual habits/rituals?

    Stephen King: I start work around 8AM and usually finish around noon. If there’s more to do, I do it in the late afternoon, although that isn’t prime time for me. The only ritual is making tea. I use the loose leaves and drink it by the gallon.

  • Paulo Coelho (Goodreads interview): late evening? (?-4AM)

    Goodreads: Do you write as soon as you wake up in the morning?

    Paulo Coelho: First I say that I’m going to write as soon as I wake up. Then I postpone and postpone and start feeling guilty and horrible and feel that I don’t deserve anything. Then I say, OK, today I’m not going to write. Then I write just to not feel guilty, and I’m going to write the first sentence. Then once I’m off the ground, the plane takes off…when I’m writing, I wake up around 12 o’clock because I write until 4 in the morning. Only two weeks. Then of course, I have to make the corrections and do another draft. I have to correct the second draft. So the first draft has, let’s say, one-third more pages than the final draft. So I start cutting.

  • Brandon Sanderson (FAQ): afternoon+late evening (noon-5PM, 8PM-3AM)

    When I was in college, I got a job working the graveyard shift at a hotel, which was great for my writing because I was there most weeknights from 11 pm until 7 am, and the only requirements that they put me to were, Just don’t fall asleep. Do whatever you want, just don’t fall asleep. We need you awake in case there’s an emergency or if anyone comes in. I ended up spending a lot of my time working on novels during those early morning hours, and that’s how I was able to pay for school, attend it full-time, and still have time for writing. I still do most of my writing in the middle of the night…

    2012 chat interview:

    Sanderson works best at night. I get up about noon, he says, write until five, and then spend a few hours with the family before starting work again about eight o’clock and then I write until the early morning hours. I often don’t get to bed before three am. He did try getting up at what most people would consider to be a more normal hour, but after a few weeks his wife capitulated, saying, This routine is making you miserable. Go back to being a night owl!

  • Margaret Atwood (Goodreads interview): morning-afternoon

    Goodreads: Can you describe a typical day spent writing?

    Margaret Atwood: There are no typical days spent writing. Let’s pretend there is one. I would get up. We would have breakfast. Then we have the coffee. That is something I really like to have to get myself started. Then I would probably sit down and type something that I had written in manuscript the day before. It’s a kind of overlap method, in which I’m typing out what I did the day before to get myself going for what I’m going to add on to that. I’m revising and then continuing to write in the same day. Then I do the next bit of new writing in the afternoon. I don’t go by how much time I spent at it but how many pages I managed to complete.

    Goodreads: Would you say you have any unusual writing habits?

    Margaret Atwood: I’m not particularly obsessive about that. But I don’t like other people using my computer. Who does like that?

  • Saint-Exupéry, in writing The Little Prince preferred evenings, starting at 11PM, or even later, as described by Lewis Galantière in 1947

    Saint-Exupéry wrote beautifully, but at the price of great effort. He went out rarely, but he had friends in almost every day to lunch and dinner. In the evening, when his friends had gone, he would brew himself a great pot of coffee and sit down to work at his dining table (his desk served merely as a catchall in which his checkbook could never be found). Now and then he would write in an all-night restaurant, where, having eaten a dish of raw chopped beef drowned in olive oil and crusted with pepper, be was likely to scribble from 2 in the morning until dawn. When be had written himself stiff, be would stretch out at home on a sofa under a lamp, take up the mouthpiece of a dictaphone, and record his copy, revising as he went along. Then, towards 7 or 8 o’clock in the morning, he would go to bed. The secretary would come in at 9 and type while be slept. Often, when friends arrived for lunch at 1 o’clock, they would ring and pound for 20 minutes before he woke up and let them in.

  • Neil Gaiman is described as a night owl (eg collaborators interviewed in Conversations With the Dream King describe him as phoning them usually late at night or as a vampire)
  • John Irving (Paris Review interview) appears to write mostly in the morning to afternoon:

    A novel is such a long involvement; when I’m beginning a book, I can’t work more than two or three hours a day. I don’t know more than two or three hours a day about a new novel. Then there’s the middle of a book. I can work eight, nine, twelve hours then, seven days a week - if my children let me; they usually don’t. One luxury of making enough money to support myself as a writer is that I can afford to have those eight-, nine-, and twelve-hour days. I resented having to teach and coach, not because I disliked teaching or coaching or wrestling but because I had no time to write. Ask a doctor to be a doctor two hours a day. An eight-hour day at the typewriter is easy; and two hours of reading over material in the evening, too. That’s routine. Then when the time to finish the book comes, it’s back to those two- and three-hour days. Finishing, like beginning, is more careful work.

  • Donald Hall

    Back then, I wrote all day, getting up at five. By this time, I rise scratchy at six or twitch in bed until seven. I drink coffee before I pick up a pen. I look through the newspaper. I try to write all morning, but exhaustion shuts me down by ten o’clock.

  • Hunter Thompson: Carroll’s biography includes a supposed daily schedule where Thompson starts writing at midnight until 6AM (but the schedule is so clearly exaggerated & humorous in the level of drug use claimed that I don’t know how seriously to take any of it); one of his editors, Terry McDonnell says When he got you on the phone in the middle of the night to listen to someone in his kitchen read to you what he had just written, all you could say was that it sounded good and that he should send it to you; a collection of letters includes him telling a landlady to put down carpets so his typing late at night won’t keep her up; his Paris Review interviewer describes him as keeping late hours on Owl Farm and the interview went into the night, where he describes his first writing job as having the advantage of letting him write entirely at night. All together, there’s no doubt Thompson preferred late night, and midnight-6AM specifically does seem plausible.
  • Michel Houellebecq (2018 roundtable): The author does not have any special routine for writing, he just likes to start working early in the morning.

TODO

  • 268 interviews, might be worth reading them all: https://www.goodreads.com/featured_lists/1-author-interviews
  • Daily Rituals: How Artists Work, Currey 2013
  • https://www.brainpickings.org/2012/11/20/daily-routines-writers/
  • The Miracle Morning for Writers: How to Build a Writing Ritual That Increases Your Impact and Your Income, Elrod; http://jamesclear.com/daily-routines-writers
  • http://www.slate.com/articles/arts/culturebox/features/2013/daily_rituals/daily_rituals_is_waking_up_early_the_secret_to_artistic_success.html
  • https://www.amazon.com/Odd-Type-Writers-Obsessive-Techniques/dp/0399159940/
  • When: The Scientific Secrets of Perfect Timing, Pink

  1. We should note that considerable doubt has been cast on both ego depletion and decision fatigue as part of the reproducibility crisis.

  2. Having been involved with modafinil for a long time, I can say that in the anecdotes I have read or talking to people or in my modafinil survey, I have seen many people praise it for nonfiction writing or programming, but I am not sure I have ever seen someone describe trying to write fiction on it, and a not infrequent complaint is that people feel less creative on it.

  3. With the striking exceptions of Balzac’s love of coffee and Beat Generation writers like Jack Kerouac or William Burroughs, who, however, in a case of the exception proving the rule, tended to write semi-autobiographical fiction, and heavily indulged in amphetamines for sleep deprivation to the point of amphetamine psychosis, assisted by wide availability at the time due to asthma treatment and diet prescriptions, see Rasmussen 2008, On Speed.

  4. Proportion/binomial test of responses split by chronotype:

    prop.test(matrix(c(196,193,467,633), nrow=2))
    #
    #   2-sample test for equality of proportions with continuity correction
    #
    # data:  matrix(c(196, 193, 467, 633), nrow = 2)
    # X-squared = 7.0006564, df = 1, p-value = 0.008147983
    # alternative hypothesis: two.sided
    # 95% confidence interval:
    #  0.01545214897 0.10848738773
    # sample estimates:
    #       prop 1       prop 2
    # 0.2956259427 0.2336561743
    df <- as.data.frame(matrix(c(196,193,467,633), nrow=2))
    df$V3 <- df$V1+df$V2
    df$Morning <- c(TRUE, FALSE)
    library(brms)
    brm(V1 | trials(V3) ~ Morning, family=binomial(), data=df, chains=32)
    # ...Population-Level Effects:
    #             Estimate Est.Error l-95% CI u-95% CI Eff.Sample Rhat
    # Intercept      -1.19      0.08    -1.35    -1.03      28662 1.00
    # MorningTRUE     0.32      0.12     0.09     0.55      26242 1.00