Skip to content
April 20, 2014 / Damien Irving

Authorea: the future of scientific writing?

It’s fair to say that LaTeX has gained widespread acceptance as the tool of choice for writing scientific scholarly articles (if you need convincing, see here, here and here). In comparison to a typical “what you see is what you get” (WYSIWYG) editor like Microsoft Word or Apache OpenOffice, the most radical aspect of LaTeX is that you don’t immediately see how your document will be typeset. Documents are instead prepared by writing a plain text input file that includes markup commands to specify the formatting, before invoking the LaTeX program to (typically) generate a final PDF document.

Since the LaTeX software and associated text editors like Texmaker are free to download, most scientists do all their document preparation on their own computer. While this is a perfectly valid workflow, it fails to take advantage of the fact that we now live in a web-enabled and highly interconnected world. As Alberto Pepe noted in his presentation at the I Annotate 2013 conference in San Fransisco, today’s scientists are doing 21st century science, writing up using 20th century writing tools (e.g. Microsoft Word, LaTeX), then locking that text away in a 17th century format (i.e. the PDF of a journal paper today has much the same format and accessibility as a scanned copy of a journal article from hundreds of years ago).

In an attempt to bring scientific writing into the 21st century, a number of online LaTeX editors have begun to appear in recent years. An obvious advantage to online editing is that you don’t need to have LaTeX installed on your machine, however since installation is both free and relatively straightforward, this hardly represents a compelling reason for scientists to change the way they write. Instead, it’s the opportunity for collaboration and sharing that has the science community so excited about online LaTeX editing.

Online LaTeX editing is an area that many people are trying to innovate in at the moment (I came across a number of “sorry we’re shutting down because we couldn’t make any money” posts in researching this article), however the two editors that have gained the most traction are ShareLaTeX and writeLaTeX. In a nutshell, these editors are to LaTeX what Google Docs is to WYSIWYG editors like Word and OpenOffice. External collaborators can view and edit the document and there are comment and chat features for discussing changes. They also provide a kind of WYSIWYG functionality, as the PDF output can be generated alongside the text editor as you type. While this is an exciting step forward, the end result for a document on ShareLaTeX and writeLaTeX is still a PDF that locks away the text in that familiar 17th century (and not to mention proprietary) format. To fully exploit the advantages of the web, a different model is clearly needed.

Alberto Pepe and his co-founder Nathan Jenkins might just be on the way to establishing that new 21st century model for scientific writing and publishing. At the most basic level, their new website Authorea offers most* of the features that ShareLaTeX and writeLaTeX do in terms of collaborative editing and commenting. In fact, their referencing (just put in a DOI and it figures out the rest), backup (you can link to your GitHub account** instead of just DropBox), PDF / Microsoft Word export (pick a journal and it will format accordingly) and IPython notebook (you can include the code that was used to generate your figures – see here for an example) functionality represents a step forward on their competitors. At a higher level, Authorea provides a way forward from that 17th century publishing format. While it does allow authors to export to PDF (or Word), the most novel thing about Authorea is that it complies your LaTeX text (or Markdown, which may be the future of scholarly writing – see here) to HTML. Instead locking your text away in a proprietary format, Authorea makes it available on the web for anyone to view and comment on.

One of the most exciting advances in scientific publishing in the last few years has been the rise of pre-print servers like arXiv. Since there is such a long time-lapse between when authors submit a manuscript and when the work is finally published, scientists are posting their work (in PDF format) to a pre-print server as soon as it’s done, so that the wider community can read their work while it’s being reviewed. In a recent interview, Nathan revealed that his “crazy long-term dream” is to fundamentally change the way that pre-publishing works. His hope is that people will pre-publish with Authorea instead, whereby their article is available in HTML format and people can comment directly on different sections of the text. This model is essentially a mix between old-school PDF journal articles and new-age blog posts, which is why there’s so much buzz around Authorea at the moment (e.g. see reviews on AppStorm and AppVita).

I recently created an account for my PhD thesis, so why not give Authorea a try for your next journal paper or manuscript?

*It should be noted that with writeLaTeX and ShareLaTeX you basically get complete access to all of the markup commands and packages that LaTeX offers. Authorea makes some minor compromises in order to compile to HTML, which are outlined in this handy cheat sheet. In other words, they cater to the needs of the vast majority of LaTeX users, but if you’re in the advanced minority of users who require highly specific figure or output formatting options then writeLaTeX or ShareLaTeX might be a better option.

**ShareLaTeX has also just introduced a way to link with GitHub, however GitHiub is the only option. In principle, you can sync Authorea with any external hosting service that is compatible with git (e.g GitHub, BitBucket, SourceForge, etc)



Leave a Comment
  1. Fiona / Apr 22 2014 14:30

    Having just begun using writeLatex this might be the push I needed to take that further step and try out Authorea. Thanks!

    • DanielC / Mar 28 2015 18:43

      I tried Authorea first, and honestly, I think it sucks. The worst part is that it is unreliable. I have written extremely basic documents (a few paragraphs, one small table, two figures) that simply failed to export to PDF. The rendering is broken. I wanted my LaTeX table to have a horizontal line and a vertical line, and Authorea could not do it. I’m sorry, but I don’t think that a line on a table is an “advanced feature”. I also found it irritating and unnatural to be constantly breaking my work into a lot of small chunks, each one with a title. For example, the moment you insert a figure you are forced to break a block.

      In the end, I was trying to like Authorea but I could not honestly recommend it to anyone. That’s when I googled for alternatives to Authorea. I heard about writeLaTeX, which has re-launched as Overleaf (and added a “rich text editor” as an alternative to straight LaTeX). I quickly fell in love with Overleaf. It actually works, and works well. You and your coworker actually get to edit the document at the same time without being forced to divide it into chunks like Authorea. Very importantly, the PDF export is extremely reliable because Overleaf is actually running LaTeX in the server and you have access to the entire LaTeX source code if you want to have it.

      • Damien Irving / Mar 29 2015 20:21

        Hi Daniel,

        I totally agree that the export bugs and table restrictions are super annoying. I guess I would say that I (and many other Authorea users I know) are willing to put up with this early adopter pain because we can see how great the concept is. Authorea is basically like a blog for your academic papers where you can write in LaTeX. If all that mattered to me was LaTeX functionality then I probably would just use Overleaf, but I can see the open science / sharing potential in the blog-like aspects of Authorea (which you don’t get with Overleaf) so I’m willing to stick with them while they iron out the bugs. Having said that, I would imagine many new users have a similar reaction to yours when confronted with the unreliable aspects of Authorea and simply never come back. It’s a big issue that they need to address pretty quickly.

      • DanielC / Mar 29 2015 22:47

        Hi Damien,

        I think you are too generous to Authorea. An annoyance is something you can work around, like the issue with the blocks. But not being able to perform the fundamental function of actually producing a paper (cause PDF output fails, or tables don’t work) is a very serious issue.

        Can you explain what’s great about the Authorea concept compared to Overleaf? I don’t see how Authorea is more or a blog, or why a blog is desirable.

  2. Administrator / May 3 2014 21:06

    Reblogged this on .

  3. Daniel Rothenberg / Sep 8 2014 22:59

    Hi Damien,

    Found your blog through your posting to the PyAOS list-serv this morning… I’ve very much enjoyed reading about your scientific workflow, and I’m certainly going to tweak my own methods with some things I think you do better! I have to comment here though. I’ve been trying to craft together a paper for a journal article in Authorea, and my impression is very simple:

    You can’t write an article on the Authorea platform.

    It simply doesn’t have the necessary tools for drafting, mind-mapping, and sketching ideas. Of course, those aren’t totally necessary, but what kills Authorea for the writing phase of a project is its cumbersome workflow. Dragging/dropping sources from Mendeley doesn’t really work as smoothly as it could (for instance, it refuses to accept my naming convention for citation references, which is a huge annoyance), and the whole “entire document on one giant freaking page” makes it hard to work things out.

    However, I think Authorea *nails* everything else. So my current article-writing project is going to incorporate Authora in this way: I’m drafting things on my desktop, and then I’ll import everything into Authorea to share my draft with my advisor and co-authors. Authorea will be great for iterative refinement, just not document creation in the first place.

    • Damien Irving / Sep 9 2014 08:18

      Hey Daniel,

      I agree – drafting is a little difficult in Authorea because you’re locked into that format of discrete text boxes. If you think of a way they could improve that aspect of the site then do let them know – I’ve found the administrators to be very responsive. Having said that, it’s difficult to think of an easy solution…

      • DanielC / Mar 28 2015 18:47

        The way to improve the “you’re locked into discrete text boxes” is to not have it. It is not needed. Google docs doesn’t have it. Overleaf doesn’t have it. Two authors can just write at the same time. So why is it that Authorea cannot exist without it?

  4. Damien Irving / Sep 21 2014 15:57

    Here’s a nice case study that goes the other direction – it links an already existing LaTeX document on GitHub to Authorea.

  5. ManuscriptPro / Jan 29 2015 13:39

    While LaTeX is a new format, not everyone is going to be able to hope on board.

    We have realized that many authors are more comfortable utilizing the basics, particularly Microsoft Word (as it easily allows change tracking).

    • DanielC / Mar 28 2015 18:51

      What? LaTeX is new? The initial release of LaTeX was in 1984. The initial release of TeX was in 1978. TeX appeared before the first version of Word for MS-DOS. How do you get to call LaTeX a “new” format?

  6. wif / May 3 2015 21:05

    I do not understand if with Authorea everybody could see your paper in progress.
    Is it not dangerous?

    • Damien Irving / May 12 2015 10:19

      As it stands at the moment ( all users get one free private article, however yes, the idea is that people can see your paper in progress. Your fear of being “scooped” (i.e. others seeing your work and publishing it before you do) is one shared by a lot of people, however many would argue that having nobody notice your work is an even bigger risk (e.g. Having your work immediately and openly accessible on Authorea (as opposed to behind a journal paywall after many months of delays in peer review and publishing) makes it easier for people to find and build on your work, which will hopefully lead to more collaborations and citations!

    • Huinca / May 20 2015 01:27

      I think it’s a terrible idea! Also, the interface is quite clumsy too. I like innovative things, but authorea to me feels quite half-baked to be honest.

      • danielcarrera / May 28 2015 16:43

        I agree. Authorea definitely seems half-baked and very poorly done. The UI is clunky for no good reason. There are plenty of other online collaborative tools that actually allow collaboration and do not force you to split your text into a bunch of little boxes. The fact that a simple PDF export of a simple document can completely fail in irreparable ways is a strong indication that the people behind Authorea don’t really know what they are doing.


  1. What’s in your bag? | Dr Climate
  2. What’s in your bag? | Dr Climate

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: