Have you heard of NaNoWriMo? National Novel Writing Month has been around since 1999 and is now far more widespread than national boundaries and has become relatively large, with 325,142 participants on six continents in 2014. The idea is simple: over the 30 days of November, you write 50,000 words that are (notionally) all directed towards a fictional novel.
I’ve taken part twice in the past and produced two … rapidly written works of fiction. I have never claimed to be a good writer, I’m certainly not a published writer, but I can put a number of words down on the page in a day. They even make sense, most of the time, and I’ve ended up with stories that real people have actually read and enjoyed!
I like NaNoWriMo. I like it as a concept, because it demystifies the concept of writing that many words by saying “Hey! Don’t get caught up on perfect prose, just start by writing.” I like the community because, despite the large number of people who show up and have no intention of doing it, there are enough like minds to give you support when you need it. I like it personally, as it’s a great way to get down a long draft of a work, even if you don’t do anything else with it. It’s a bit of external structure (and scaffolding) for those of us who aren’t professional authors.
Being me, of course, I’m all about seeing if we can get people doing something that they didn’t think possible, so I look at NaNoWriMo as a success for anyone who writes one more word than they otherwise would have in November. Sure, getting down a whole novel would be awesome but any steps forward are good steps.
We could talk a lot about how this kind of constrained activity can work in a creative setting but, if you know my work, you’ll have a fairly good idea that I think that everyone taking part should have “enjoyment” as their primary goal, with a possible outcome of their first long-form work as a happy side-effect. 50K shouldn’t be a burden but a guide. 1,700 words a day can be more manageable than many people think and, at the end of November, you may have done something you never thought possible.
Whatever happens, you’ll be thinking creatively and that, in my book, is always awesome. “Almighty creativity”, as the late Bob Ross might say.
I’ll be doing NaNo again this year and, if you’re thinking about it, check out the web site or my shortish guide to speed writing (AntifreezePub) that I’ve written based on my own experiences over the years. As always, if you think there should be a better speed writing guide, feel free to write it/find it and link to it in the comments!
I decided to have an outrageous book cover to inspire me and get me into the right mood. (Currently a working copy with placeholder artwork, as I have no idea what will make it into the final draft.)
Almost all of us benefit from writing practice and this is an interesting way to get a lot of practice in a short time. If you do it, have fun, and feel free to buddy up with me under the username jnick.
Let me start by putting up a picture of some people celebrating!
My first confession is that the ‘acceptance’ I’m talking about is for academic and traditional fiction publishing. The second confession is that I have attempted to manipulate you into clicking through by using a carefully chosen title and presented image. This is to lead off with the point I wish to make today: we are a mess of implicit and explicit cognitive biases and to assume that we have anything approximating a fair evaluation mechanism to get work published is to, sadly, be making a far reaching assumption.
If you’ve read this far, my simple takeaway is “If people don’t even start reading your work with a positive frame of mind and a full stomach, your chances of being accepted are dire.”
If you want to hang around my argument is going to be simple. I’m going to demonstrate that, for much simpler assessments than research papers or stories, simple cognitive biases have a strong effect. I’m going to follow this and indicate how something as simple as how hungry you are can affect your decision making. I’m then going to identify a difference between scientific publishing and non-scientific publishing in terms of feedback and why expecting that we will continue to get good results from both approaches is probably too optimistic. I am going to make some proposals as to how we might start thinking about a fix, but only to start discussion because my expertise in non-academic publishing is not all that deep and limited by not being an editor or publisher!
[Full disclosure: I am happily published in academia but I am yet to be accepted for publication in non-academic approaches. I am perfectly comfortable with this so please don’t read sour grapes into this argument. As you’ll see, with the approaches I propose, I would in fact strip myself of some potential bias privileges!]
I’ve posted before on an experiment  where the only change to the qualifications of a prospective lab manager was to take the name from male to female. The ‘female’ version of this CV got offered less money, less development support and was ‘obviously’ less qualified. And this effect occurred whether the assessor was a man or a woman. This is the pretty much the gold standard for experiments of this type because it reduced any possibility of someone acting out of character because they knew what the experiment was trying to prove. There’s a lot of discussion in fiction at the moment about gendered bias, as well as academia. You’re probably aware of the Bechdel Test, which simply asks if there are two named women in a film who talk to each other about something other than men, and how often the mainstream media fails that test. But let’s look at something else. Antony LaPaglia tells a story that he used to get pulled up on his American accent whenever anyone knew that he was Australian. So he started passing as American. Overnight, complaints about his accent went away.
Compared to assessing a manuscript, reading a CV, bothering to put in two woman with names and a story, and spotting an accent are trivial and yet we can’t get these right without bias.
There’s another thing called the Matthew Effect, which basically says that the more you have, the more you’re going to get (terrible paraphrasing). Thus, the first paper in a field will be one of the most cited, people are comfortable giving opportunities to people who have used them well before, and so on. It even shows up in graph theory, where the first group of things connected together tend to become the most connected!
So, we have lots of examples of bias that comes in, if we know enough about someone that the bias can engage. And, for most people who aren’t trying to be discriminatory, it’s actually completely unconscious. Really? You don’t think you’d notice?
Let’s look at the hunger argument. An incredible study  (Economist link for summary) shows that Israeli judges are less likely to grant parole, the longer they’ve waited since they ate, even when taking other factors into account. Here’s a graph. Those big dips are meal breaks.
When confronted with that terrifying graph, the judges were totally unaware of it. The people in the court every day hadn’t noticed it. The authors of the study looked at a large number of factors and found some things that you’d expect in terms of sentencing but the meal break plunges surprised everyone because they had never thought to look for it. The good news is that, most days, the most deserving will still get paroled but, and it’s a big but, you still have to wonder about the people who should have been given parole who were denied because of timing and also the people who were paroled who maybe should not have been.
So what distinguishes academia and non-academic publishing? Shall we start by saying that, notionally, many parts of academic publishing subscribe to the Popperian model of development where we expose ideas to our colleagues and they tear at them like deranged wolves until we fashion truth? As part of that, we expect to get reviews from almost all submissions, whether accepted or not, because that is how we build up academic consensus and find out new things. Actual publication allows you to put your work out to everyone else where they can read it, work with it or use it to fashion a counter-claim.
In non-academic publishing, the publisher wants something that is saleable in the target market and the author wants to provide this. The author probably also wants to make some very important statements about truth, beauty, the lizard people or anything else (much as in academic publishing, the spread of ideas is crucial). However, from a publisher’s perspective, they are not after peer-verified work of sufficient truth, they are after something that matches their needs in order to publish it, most likely for profit.
Both are directly or indirectly prestige markers and often have some form of financial rewards, as well as some truth/knowledge construction function. Non-academic authors publish to eat, academic authors publish to keep their jobs or get tenure (often enough to allow you to eat). But the key difference is the way that feedback is given because an academic journal that gave no feedback would have trouble staying in business (unless it had incredible acceptance already, see Matthew Effect) because we’re all notionally building knowledge. But “no feedback” is the default in other publishing.
When I get feedback academically, I can quickly work out several things:
- Is the reviewer actually qualified to review my work? If someone doesn’t have the right background, they start saying things like surely when they mean I don’t know, and it quickly tells you that this review will be uninformative.
- Has the reviewer actually read the work? I would ask all the academics reading this to send me $1 if they’ve ever been told to include something that is obviously in the paper and takes up 1-2 pages already, except I am scared of the tax and weight implications.
- How the feedback can be useful. Good feedback is great. It spots holes, it reinforces bridges, it suggests new directions.
- If I want to publish in that venue again. If someone can’t organise their reviewers and oversee the reviews properly? I’m not going to get what I need to do good work. I should go and publish elsewhere.
My current exposure to non-academic publishing has been: submit story, wait, get rejection. Feedback? “Not suitable for us but thank you for your interest”, “not quite right for us”,”I’m going to pass on this”. I should note that the editors have all been very nice, timely (scarily so, in some cases) and all of my interactions have been great – my problem is mechanistic, not personal. I should clearly state that I assume that point 1 from above holds for all non-academic publishing, that is that the editors have chosen someone to review in a genre that they don’t actually hate and know something about. So 1 is fine. But 2 is tricky when you get no feedback.
But that tricky #2, “Has the reviewer actually read the work”, in the context of my previous statements really becomes “HOW has the reviewer read my work?” Is there an informal ordering of people you think you’ll enjoy to newbies, even unconsciously? How hungry is the reviewer when they’re working? Do they clear up ‘simple checks’ just before lunch? In the absence of feedback, I can’t assess the validity of the mechanism. I can’t improve the work with no feedback (step 3) and I’m now torn as to whether this story was bad for a given venue or whether my writing is just so awful that I should never darken their door again! (I accept, dear reader, that this may just be the sad truth and they’re all too scared to tell me.)
Let me remind you that implicit bias is often completely unconscious and many people are deeply surprised when they discover what they have been doing. I imagine that there are a number of reviewers reading this who are quite insulted. I certainly don’t mean to offend but I will ask if you’ve sat down and collected data on your practice. If you have, I would really love to see it because I love data! But, if what you have is your memory of trying to be fair… Many people will be in denial because we all like to think we’re rational and fair decision makers. (Looks back at those studies. Umm.)
We can deal with some aspects of implicit bias by using blind review systems, where the reviewer only sees the work and we remove any clues as to who wrote it. In academia this can get hard because some people’s contributed signature is so easy to see but it is still widely used. (I imagine it’s equally hard for well known writers.) This will, at least, remove gender bias and potentially reduce the impact of “famous people”, unless they are really distinctive. I know that a blinding process isn’t happening in all of the parts of non-academic publishing because my name is all over my manuscripts. (I must note that there are places that use blind submission, such as Andromeda Spaceways Inflight Magazine and Aurealis, for initial reading, which is a great start.) Usually, when I submit, my covering letter has to clearly state my publication history. This is the very opposite of a blind process because I am being asked to rate myself for Matthew Effect scaling every time I submit!
(There are also some tips and tricks in fiction, where your rejections can be personalised, yet contain no improvement information. This is still “a better rejection” but you have to know this from elsewhere because it’s not obvious. Knowing better writers is generally the best way to get to know about this. Transparency is not high, here.)
The timing one is harder because it requires two things: multiple reviewers and a randomised reading schedule, neither of which take into account the shoe string budgets and volunteer workforce associated with much of fiction publishing. Ideally, an anonymised work gets read 2-3 times, at different times relative to meals and during the day, taking into account the schedule of the reader. Otherwise, that last manuscript you reject before rushing home at 10pm to reheat a stale bagel? It would have to be Hemingway to get accepted. And good Hemingway at that.
And I’d like to see randomised reading applied across academic publishing as well. And we keep reviewing it until we actually reach a consensus. I’ve been on a review panel recently where we had two ‘accepts’, two ‘mehs’ and two ‘kill it with fires’ for the same paper. After group discussion, we settled for ‘a weak accept/strong meh’. Why? Because the two people who had rated it right down weren’t really experts so didn’t recognise what was going on. Why were they reviewing? Because it’s part of the job. So don’t think I’m going after non-academic publishing here. I’m exposing problems in both because I want to try and fix both.
But I do recognise that the primary job of non-academic publishing is getting people to read the publication, which means targeting saleable works. Can we do this in a way that is more systematic than “I know good writing when I see it” because (a) that doesn’t scale and (b) the chances of that aligning across more than two people is tiny.
This is where technological support can be invaluable. Word counting, spell checking and primitive grammar checking are all the dominion of the machine, as is plagiarism detection on existing published works. So step one is a brick wall that says “This work has not been checked against our submissions standards: problems are…” and this need not involve a single human (unless you are trying to spellcheck The Shugenkraft of Berzxx, in which case have a tickbox for ‘Heavy use of neologisms and accents’.) Plagiarism detection is becoming more common in academic writing and it saves a lot of time because you don’t spend it reading lifted work. (I read something that was really familiar and realised someone had sent me some of my own work with their name on it. Just… no.)
What we want is to go from a flood, to a river, then to manage that river and direct it to people who can handle a stream at a time. Human beings should not be the cogs and failure points in the high volume non-academic publishing industry.
Stripping names, anonymising and randomly distributing work is fairly important if we want to remove time biases. Even the act of blinding and randomising is going to reduce the chances that the same people get the same good or bad slots. We are partially systematic. Almost everyone in the industry is overworked, doing vast and wonderful things and, in the face of that, tired and biassed behaviour becomes more likely.
The final thing that would be useful is something alone the lines of a floating set of check boxes that sit with the document, if it’s electronic. (On paper, have a separate sheet that you can scan in once it’s filled in and then automatically extract the info.) What do you actually expect? What is this work/story not giving you? Is it derivative work? Is it just all talk and no action? Is it too early and just doesn’t go anywhere? Separating documents from any form of feedback automation (or expecting people to type sentences) is going to slow things down and make it impossible to give feedback. Every publishing house has a list of things not to do, let’s start with the 10 worst of those and see how many more we can get onto the feedback screen.
I am thinking of an approach that makes feedback an associated act of reading and can then be sent, with accept or reject, in the same action. Perhaps it has already been created and is in use in fine publishing houses, but my work hasn’t hit a bar where I even get that feedback? I don’t know. I can see that distributed editorial boards, like Andromeda, are obviously taking steps down this path because they have had to get good at shunting stuff around at scale and I would love to know how far they’ve got. For me, a mag that said “We will always give you even a little bit of feedback” will probably get all of my stuff first. (Not that they want it but you get the idea.)
I understand completely that publishers are under no obligation whatsoever to do this. There is no right to feedback nor is there an expectation outside of academia. But if we want good work, then I think I’ve already shown that we are probably missing out on some of it and, by not providing feedback, some (if not many) of those stories will vanish, never worked on again, never seen again, because the authors have absolutely no guidance on how to change their work.
I have already discussed mocking up a system, building from digital humanist approaches and using our own expertise, with one of my colleagues and we hope to start working on something soon. But I’d rather build something that works for everyone and lets publishers get more good work, authors recognised when they get it right, and something that brings more and more new voices into the community. Let me know if it’s already been written or take me to school in the comments below. I can’t complain about lack of feedback and then ignore it when I get it!
 PNAS, vol. 109 no. 41, Corinne A. Moss-Racusin, 16474–16479, doi: 10.1073/pnas.1211286109
 PNAS vol. 108 no. 17, Shai Danziger, 6889–6892, doi: 10.1073/pnas.1018033108
I occasionally dump the blog and run it through some Python script deliciousness to find out how many words I’ve written. This is no measure of worth or quality, more a metric of my mania. As I noted in October, I was going to hit what I thought was my year target much earlier. Well, yes, it came and it went and, sure enough, I plowed through it. At time of writing, on published posts alone, we’re holding at around 1.2 posts/day, 834 words/post and a smidgen over 340,000 words, which puts me (in word count) just after Ayn Rand’s “The Fountainhead” (311,596) but well behind her opus “Atlas Shrugged” (561,996). In terms of Objectivism? Let’s just say that I won’t be putting any kind of animal into that particular fight at the moment.
Now, of course, I can plug in the numbers and see that this puts my final 2012 word count somewhere in the region of 362,000 words. I must admit, there is a part of me that sees that number and thinks “Well, we could make it an even 365,000 and that’s a neat 1000 words/day” but, of course, that’s dumb for several reasons:
- I have not checked in detail exactly how well my extraction software is grabbing the right bits of the text. There are hyperlinks and embellishments that appear to be taken care of, but we are probably only on the order of 95% accuracy here. Yes, I’ve inspected it and I haven’t noticed anything too bad, but there could be things slipping through. After all of this is over, I am going to drag it all together and analyse it properly but, let me be clear, just because I can give you a word count to 6 significant figures, doesn’t mean that it is accurate to 6 significant figures.
- Should I even be counting those sections of text that are quoted? I do like to put quotes in, sometimes from my own work, and this now means I’m either counting something that I didn’t write or I’m counting something that I did write twice!
- Should I be counting the stats posts themselves as they are, effectively, metacontent? This line item is almost above that again! This way madness lies!
- It was never about the numbers in the first place, it was about thinking about my job, my students, my community and learning and teaching. That goal will have been achieved whether I write one word/day from now on or ten thousand!
But, oh, the temptation to aim for that ridiculous and ultimately deceptive number. How silly but, of course, how human to look at the measurable goal rather than the inner achievement or intrinsic reward that I have gained from the thinking process, the writing, the refining of the text, the assembly of knowledge and the discussion.
Sometime after January the 1st, I will go back and set the record straight. I shall dump the blog and analyse it from here to breakfast time. I will release the data to interested (and apparently slightly odd) people if they wish. But, for now, this is not the meter that I should be watching because it is not measuring the progress that I am making, nor is it a good compass that I should follow.
I tidied up my office recently and managed to clear out about a couple of boxes full of old paper. Some of these were working drafts of research papers, covered in scrawl (usually red because it shows up more), some were book chapter mark-ups, and some were things like project meeting plans that I could scribble on as people spoke. All of this went into either the secure waste bins (sekrit stuff) or the general recycling because I do try to keep the paper footprint down. However, my question to myself is two-fold:
- Why do I still have an office full of paper when I have a desktop, (two) laptops, an iPad and an iPhone, and I happily take notes and annotate documents on them?
- Why am I surrounded by so many books, still?
I don’t think I’ve ever bought as many books as I have bought this year. By default, if I can, I buy them as the electronic and paper form so that I can read them when I travel or when I’m in the office. There are books on graphic design, books on semiotics, books on data visualisation and analysis, and now, somewhat recursively, books on the end of books. My wife found me a book called “This is not the end of the book”, which is a printed conversation between Umberto Eco and Jean-Claude Carrière, curated by Jean-Philippe de Tonnac. I am looking forward to reading it but it has to wait until some of the other books are done. I have just finished Iain M. Banks latest “The Hydrogen Sonata”, am swimming through an unauthorised biography of Led Zeppelin and am still trying to finish off the Derren Brown book that I have been reading on aeroplanes for the past month or so. Sitting behind all this are “Cloud Atlas” and “1Q84”, both of which are officially waiting until I have finished my PhD application portfolio for creative writing. (Yes, dear reader, I’m nervous because they could as easily say ‘No’ as ‘Yes’ but then I will learn how to improve and, if I can’t take that, I shouldn’t be teaching. To thine own dogfood, be as a consumer.)
Why do I still write on paper? Because it feels good. I select pens that feel good to write with, or pencils soft enough to give me a good relationship to the paper. The colour of ink changes as it hits the paper and dries and I am slightly notorious for using inks that do not dry immediately. When I was a winemaker, I used black Bic fine pens, when many other people used wet ink or even fountain pens, because the pen could write on damp paper and, even when you saturated the note, the ink didn’t run. These days, I work in an office and I have the luxury of using a fountain pen to scrawl in red or blue across documents, and I can enjoy the sensation.
Why do I still read on paper? Because it is enjoyable and I have a long relationship with the book, which began from a very early age. The book is also, nontrivially, one of the few information storage devices that can be carried on to a plane without having to be taken from one’s bag or shut down for the periods of take off and landing. I am well aware of the most dangerous points in an aircraft’s cycle and I strongly prefer to be distracted by, if not in-flight entertainment, then a good solid book. But it is also the pleasure of being able to separate the book from the devices that link me into my working world, yet without adding a new data storage management issue. Yes, I could buy a Kindle and not have to check my e-mail, but then I have to buy books from this store and I have to carry that charger or fit it next to my iPad, laptop and phone when travelling. Books, once read, can either be donated to your hosts in another place or can be tossed into the suitcase, making room for yet more books – but of course a device may carry many books. If I have no room in my bag for a book, then I don’t have to worry about the fragility of making space in my carry-on by putting it into the suitcase.
And, where necessary, the book/spider interaction causes more damage to the spider than the contents of the book. My thesis was sufficiently large to stun a small mammal, but you would not believe how hard it was to get ethical approval for that!
The short answer to both questions is that I enjoy using the physical forms although I delight in the practicality, the convenience and the principle of the electronic forms. I am a happy hybridiser who wishes only to enjoy the experience of reading and writing in a way that appeals to me. In a way, the electronic format makes it easier for me to share my physical books. I have a large library of books from when I was younger that, to my knowledge, has books that it is almost impossible to find in print or libraries any more. Yet, I am in that uncomfortable position of being a selfish steward, in that I cannot release some of these books for people to read because I hold the only copy that I know of. As I discover more books in electronic or re-print format (the works of E. Nesbit, Susan Cooper in the children’s collection of my library, for example) then I am free to use the books as they were intended, as books.
What we have now, what is emerging, certainly need not be the end of the book but it will be interesting to look back, in fifty years or so, to find out what we did. If the book has become the analogue watch of information, where it moved from status symbol for its worth, to status symbol for its value, to affectation and, now, to many of my students, an anachronism for those who don’t have good time signal on their phones. I suspect that a watch does not have the sheer enjoyability of the book or the pen on paper, but, if you will excuse me, time will tell.
My posts recently have been getting longer and longer and I think I’m hitting the point where ‘prolix’ is an eligible adjective: I’m at risk of using so many words that people may not finish or start reading, or risk being bored by the posts. Despite the fact that I write quickly, it does take some time to write 2,000 words. I want to write the number of words to carry the point across and make the best of your time and my time.
I’m going to experiment with posts that are as informative/useful but that are slightly shorter, aiming for 1,000 words as an upper bound and splitting posts thematically where possible to keep to this. At the end of July, assuming I remember, I’m going to review this to see how it’s going. (The risk, of course, is that editing to keep inside this frame will consume far more time than just writing. Believe me, I’m aware of that one!)
As always, feedback is very welcome and I reserve the right to completely forget about this and start writing 10,000 word megaposts again because I’ve become carried away. Thanks for reading!