James Fallows’ post about the writing program Scrivener “suggests broader truths about the ways computers help and hinder the way we think.” He’s right, although I’ve used Scrivener and didn’t love it enough to switch: for anything beyond blog posts I mostly use a combination of Microsoft Word and Mellel, a word processor that is very fast and stable but can’t track changes. This, for me, is not merely bad: I can’t use Mellel beyond first drafts.

The other problem with Mellel isn’t related to the program itself, but to the release cycle. It’s discouraging when a forum post from the developer says, “Yes, we have been slacking off. The pace of development of Mellel – that is, the number of new releases – have dropped significantly over the last three years.” That’s another way of saying, “We’re not really working on it.”

Word, in turn, gets used for any documents I have to share with others (since they already have Word).

Fallows describes how Scrivener offers “a ‘project’ organization system that makes it easy to amass many notes, files, quotes, research documents, etc related to the essay or article or book you’re writing.” I primarily use Devonthink Pro (DTP) for this kind of purpose, and it connects whatever ideas I have to other quotes, ideas, and the like. The “artificial intelligence” engine is surprisingly useful at making connections that I didn’t realize I had. Obviously I could use DTP with Scrivener, but the use of DTP makes the marginal value of Scrivener somewhat lower.

Scrivener 2.0, however, is intriguing; these videos demonstrate its power. More on that later, as I’d like to follow-up on the idea that computers can “help and hinder the way we work.” Scrivener enables one to rearrange large chunks of materials easily, which is how a lot of writers work in the off-line world. For example, I’ve been reading Critical Essays on James Joyce’s Ulysses for a seminar paper and came across this description of Joyce’s process in A. Walton Litz’s “The Design of Ulysses:”

[Joyce] did not write Ulysses straight through, following the final order the episodes. First it was necessary to determine the design of the novel, to visualize its characters and the course of the action, and this entailed putting scattered portions on paper in order to clarify them. Then, like the mosaic worker, Joyce collected and sorted material to fit the design. Finally, the fragments were placed in their proper positions through a process of rough drafts and revisions.

The “design” and the ability to “visualize its characters and the course of the action” corresponds roughly to Scrivener’s idea pane. The “scattered portions on paper” come next so they can be rearranged, “collected” and “sorted.” There’s nothing wrong with using pieces of paper, of course—it worked for Joyce!—but I wonder what the great novelist would think of working digitally.

Joyce used notecards, and Litz liked the mosaic-worker analogy so much that he uses it again a few pages later:

It was the function of the note-sheets to assure that patterns and relationships already visualized by Joyce reached their fore-ordained positions in the text. Like the mosaic worker, he was continuously sorting and re-grouping his raw materials, assigned each fragment to its proper place in the general design. The mechanical nature of this process emphasizes the mechanical nature of those ordering principles which give Ulysses its superficial unity [. . . ]

I used to write more like this and now I write less like this: it is often my goal to ensure that each chapter follows inexorably from the preceding chapter. The narrative threads and the desires of each character should force the novel in a particular direction. If I can rearrange the chapters relatively easily, then I feel like I’ve done something wrong. I still want “patterns and relationships” to reach conclusions, but I don’t want those conclusions “fore-ordained:” I want them to arise organically, and for them to be inevitable yet surprising. This is a difficult trick to pull off, but it means that the serial nature of the writing I do is probably less likely to be helped by the structure of Scrivener than the writing some others might do.

In the essay after Litz’s, Anthony Cronin’s “The Advent of Bloom” begins with the structure of Ulysses: “[. . .] if Ulysses can be said to have a plot, its plot is formless and does not give form to the book – it is not shaped to produce a series of dramatic sensations for purposes aesthetic or otherwise; it has no conclusion in event, only a termination in time [. . .]” If a plot “does not give form to the book,” then something must; for some writers, Scrivener might organize it and help find a way to present formlessness. The program helps one create a mosaic, but I’m not trying to create a mosaic in my work, at least right now: I’m trying to create a linear plot. So I don’t think the program will help me as much as it could.

Nonfiction books, on the other hand, might be much better with Scrivener: in my papers, I move material around much more frequently than I do in fiction. Since I haven’t written any nonfiction books, however, I can’t comment as much on those.

I suspect that large, high-resolution monitors enable programs like Scrivener: at 24″ or larger, one can have a broad enough swatch of material open to really make a (computer) desktop feel like a (physical) desktop. You can layout and rearrange items much more easily. The new 27″ iMacs in particular are appealing for this purpose, and one can now find 27″ external monitors from Dell, Apple, and others. As desktops become more like desktops, being able to visualize large amounts of information at once makes tools like Scrivener more useful.

At the moment, I’m about 80K words into a novel that I think will end up in the neighborhood of 100K – 110K words, which is a bit long for a first published work but not impossibly long. Using a 24″ iMac, I can easily have two pages of text open at a time, which is very convenient. That’s what I use for my “notes” section (miscellaneous stuff I want to remember but can’t immediately add to the main narrative) and my main window, which has the novel progressing from Chapter 1 to “### END ###.” On my second monitor, a 20″ cheapie Dell, I have an outline and character list open.

Some of those functions could be taken over by Scrivener, based on what I’ve seen in the videos. For my next novel—if there is another in the immediate future; I need to devote more time to academic writing—I’d be willing to try Scrivener long enough to know if version 2.0 is a good fit. For this one, however, the thought of changing tools in the middle of the process would be too disruptive. There’s no reason, after all, that I can’t use both Scrivener and Devonthink Pro.

James Fallows’ post about the writing program Scrivener “suggests broader truths about the ways computers help and hinder the way we think.” He’s right, although I’ve used Scrivener and didn’t love it enough to switch: for anything beyond blog posts I mostly use a combination of Microsoft Word and Mellel, a word processor that is very fast and stable but can’t track changes. This, for me, is not merely bad: I can’t use Mellel beyond first drafts.

The other problem with Mellel isn’t related to the program itself, but to the release cycle. It’s discouraging when a forum post from the developer says, “Yes, we have been slacking off. The pace of development of Mellel – that is, the number of new releases – have dropped significantly over the last three years.” That’s another way of saying, “We’re not really working on it.”

Word, in turn, gets used for any documents I have to share with others (since they already have Word).

Fallows describes how Scrivener offers “a ‘project’ organization system that makes it easy to amass many notes, files, quotes, research documents, etc related to the essay or article or book you’re writing.” I primarily use Devonthink Pro (DTP) for this kind of purpose, and it connects whatever ideas I have to other quotes, ideas, and the like. The “artificial intelligence” engine is surprisingly useful at making connections that I didn’t realize I had. Obviously I could use DTP with Scrivener, but the use of DTP makes the marginal value of Scrivener somewhat lower.

Scrivener 2.0, however, is intriguing; these videos demonstrate its power. More on that later, as I’d like to follow-up on the idea that computers can “help and hinder the way we work.” Scrivener enables one to rearrange large chunks of materials easily, which is how a lot of writers work in the off-line world. For example, I’ve been reading Critical Essays on James Joyce’s Ulysses for a seminar paper and came across this description of Joyce’s process in A. Walton Litz’s “The Design of Ulysses:”

[Joyce] did not write Ulysses straight through, following the final order the episodes. First it was necessary to determine the design of the novel, to visualize its characters and the course of the action, and this entailed putting scattered portions on paper in order to clarify them. Then, like the mosaic worker, Joyce collected and sorted material to fit the design. Finally, the fragments were placed in their proper positions through a process of rough drafts and revisions.

The “design” and the ability to “visualize its characters and the course of the action” corresponds roughly to Scrivener’s idea pane. The “scattered portions on paper” come next so they can be rearranged, “collected” and “sorted.” There’s nothing wrong with using pieces of paper, of course—it worked for Joyce!—but I wonder what the great novelist would think of working digitally.

Joyce used notecards, and Litz liked the mosaic-worker analogy so much that he uses it again a few pages later:

It was the function of the note-sheets to assure that patterns and relationships already visualized by Joyce reached their fore-ordained positions in the text. Like the mosaic worker, he was continuously sorting and re-grouping his raw materials, assigned each fragment to its proper place in the general design. The mechanical nature of this process emphasizes the mechanical nature of those ordering principles which give Ulysses its superficial unity [. . . ]

I used to write more like this and now I write less like this: it is often my goal to ensure that each chapter follows inexorably from the preceding chapter. The narrative threads and the desires of each character should force the novel in a particular direction. If I can rearrange the chapters relatively easily, then I feel like I’ve done something wrong. I still want “patterns and relationships” to reach conclusions, but I don’t want those conclusions “fore-ordained:” I want them to arise organically, and for them to be inevitable yet surprising. This is a difficult trick to pull off, but it means that the serial nature of the writing I do is probably less likely to be helped by the structure of Scrivener than the writing some others might do.

In the essay after Litz’s, Anthony Cronin’s “The Advent of Bloom” begins with the structure of Ulysses: “[. . .] if Ulysses can be said to have a plot, its plot is formless and does not give form to the book – it is not shaped to produce a series of dramatic sensations for purposes aesthetic or otherwise; it has no conclusion in event, only a termination in time [. . .]” If a plot “does not give form to the book,” then something must; for some writers, Scrivener might organize it and help find a way to present formlessness. The program helps one create a mosaic, but I’m not trying to create a mosaic in my work, at least right now: I’m trying to create a linear plot. So I don’t think the program will help me as much as it could.

Nonfiction books, on the other hand, might be much better with Scrivener: in my papers, I move material around much more frequently than I do in fiction. Since I haven’t written any nonfiction books, however, I can’t comment as much on those.

I suspect that large, high-resolution monitors enable programs like Scrivener: at 24″ or larger, one can have a broad enough swatch of material open to really make a (computer) desktop feel like a (physical) desktop. You can layout and rearrange items much more easily. The new 27″ iMacs in particular are appealing for this purpose, and one can now find 27″ external monitors from Dell, Apple, and others. As desktops become more like desktops, being able to visualize large amounts of information at once makes tools like Scrivener more useful.

At the moment, I’m about 80K words into a novel that I think will end up in the neighborhood of 100K – 110K words, which is a bit long for a first published work but not impossibly long. Using a 24″ iMac, I can easily have two pages of text open at a time, which is very convenient. That’s what I use for my “notes” section (miscellaneous stuff I want to remember but can’t immediately add to the main narrative) and my main window, which has the novel progressing from Chapter 1 to “### END ###.” On my second monitor, a 20″ cheapie Dell, I have an outline and character list open.

Some of those functions could be taken over by Scrivener, based on what I’ve seen in the videos. For my next novel—if there is another in the immediate future; I need to devote more time to academic writing—I’d be willing to try Scrivener long enough to know if version 2.0 is a good fit. For this one, however, the thought of changing tools in the middle of the process would be too disruptive. There’s no reason, after all, that I can’t use both Scrivener and Devonthink Pro.